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 2023/02/08 17:36:02 UTC
[tvm-site] branch asf-site updated: deploying docs (apache/tvm@2e30e772a7395afa6674235e22943c433a1a3f32)
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 87c1ab488e deploying docs (apache/tvm@2e30e772a7395afa6674235e22943c433a1a3f32)
87c1ab488e is described below
commit 87c1ab488e1cbea201d4bd4b993cd50dccb94a6d
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Wed Feb 8 17:35:55 2023 +0000
deploying docs (apache/tvm@2e30e772a7395afa6674235e22943c433a1a3f32)
---
docs/_images/sphx_glr_micro_train_001.png | Bin 337291 -> 341097 bytes
docs/_images/sphx_glr_micro_train_thumb.png | Bin 24331 -> 24317 bytes
.../how_to/compile_models/from_darknet.rst.txt | 2 +-
.../how_to/compile_models/from_keras.rst.txt | 2 +-
.../how_to/compile_models/from_mxnet.rst.txt | 2 +-
.../how_to/compile_models/from_oneflow.rst.txt | 2 +-
.../how_to/compile_models/from_pytorch.rst.txt | 2 +-
.../how_to/compile_models/from_tensorflow.rst.txt | 2 +-
.../compile_models/sg_execution_times.rst.txt | 22 +-
.../deploy_models/deploy_model_on_adreno.rst.txt | 2 +-
.../deploy_models/deploy_model_on_android.rst.txt | 2 +-
.../deploy_object_detection_pytorch.rst.txt | 4 +-
.../deploy_models/deploy_prequantized.rst.txt | 6 +-
.../deploy_prequantized_tflite.rst.txt | 4 +-
.../how_to/deploy_models/deploy_quantized.rst.txt | 2 +-
.../deploy_models/deploy_ssd_gluoncv.rst.txt | 4 +-
.../deploy_models/sg_execution_times.rst.txt | 20 +-
.../extend_tvm/bring_your_own_datatypes.rst.txt | 2 +-
.../how_to/extend_tvm/sg_execution_times.rst.txt | 8 +-
.../how_to/extend_tvm/use_pass_instrument.rst.txt | 16 +-
.../optimize_operators/opt_conv_cuda.rst.txt | 2 +-
.../optimize_operators/opt_conv_tensorcore.rst.txt | 2 +-
.../how_to/optimize_operators/opt_gemm.rst.txt | 16 +-
.../optimize_operators/sg_execution_times.rst.txt | 8 +-
.../sg_execution_times.rst.txt | 14 +-
.../tune_conv2d_layer_cuda.rst.txt | 1824 ++++++++++++++++----
.../tune_network_cuda.rst.txt | 4 +-
.../tune_network_x86.rst.txt | 4 +-
.../tune_sparse_x86.rst.txt | 481 +++++-
.../tune_with_autotvm/sg_execution_times.rst.txt | 10 +-
.../tune_with_autotvm/tune_conv2d_cuda.rst.txt | 340 +---
.../work_with_microtvm/micro_autotune.rst.txt | 16 +-
.../work_with_microtvm/micro_pytorch.rst.txt | 4 +-
.../how_to/work_with_microtvm/micro_train.rst.txt | 18 +-
.../work_with_microtvm/sg_execution_times.rst.txt | 12 +-
.../work_with_relay/sg_execution_times.rst.txt | 8 +-
.../how_to/work_with_schedules/intrin_math.rst.txt | 2 +-
.../work_with_schedules/sg_execution_times.rst.txt | 14 +-
.../how_to/work_with_schedules/tensorize.rst.txt | 2 +-
.../tutorials/autotvm/sg_execution_times.rst.txt | 6 +-
.../frontend/deploy_classification.rst.txt | 2 +-
.../tutorials/frontend/deploy_detection.rst.txt | 2 +-
.../tutorials/frontend/sg_execution_times.rst.txt | 6 +-
.../tutorials/optimize/sg_execution_times.rst.txt | 6 +-
.../topic/vta/tutorials/sg_execution_times.rst.txt | 4 +-
.../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 | 59 +-
.../tutorial/cross_compilation_and_rpc.rst.txt | 2 +-
docs/_sources/tutorial/intro_topi.rst.txt | 2 +-
docs/_sources/tutorial/sg_execution_times.rst.txt | 22 +-
.../tutorial/tensor_expr_get_started.rst.txt | 42 +-
docs/commit_hash | 2 +-
docs/how_to/compile_models/from_darknet.html | 2 +-
docs/how_to/compile_models/from_keras.html | 2 +-
docs/how_to/compile_models/from_mxnet.html | 2 +-
docs/how_to/compile_models/from_oneflow.html | 16 +-
docs/how_to/compile_models/from_pytorch.html | 13 +-
docs/how_to/compile_models/from_tensorflow.html | 2 +-
docs/how_to/compile_models/sg_execution_times.html | 26 +-
.../deploy_models/deploy_model_on_adreno.html | 2 +-
.../deploy_models/deploy_model_on_android.html | 2 +-
.../deploy_object_detection_pytorch.html | 55 +-
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 | 37 +-
docs/how_to/deploy_models/sg_execution_times.html | 20 +-
.../extend_tvm/bring_your_own_datatypes.html | 2 +-
docs/how_to/extend_tvm/sg_execution_times.html | 8 +-
docs/how_to/extend_tvm/use_pass_instrument.html | 16 +-
docs/how_to/optimize_operators/opt_conv_cuda.html | 2 +-
.../optimize_operators/opt_conv_tensorcore.html | 2 +-
docs/how_to/optimize_operators/opt_gemm.html | 16 +-
.../optimize_operators/sg_execution_times.html | 8 +-
.../sg_execution_times.html | 14 +-
.../tune_conv2d_layer_cuda.html | 1824 ++++++++++++++++----
.../tune_with_autoscheduler/tune_network_cuda.html | 4 +-
.../tune_with_autoscheduler/tune_network_x86.html | 4 +-
.../tune_with_autoscheduler/tune_sparse_x86.html | 481 +++++-
.../tune_with_autotvm/sg_execution_times.html | 10 +-
.../how_to/tune_with_autotvm/tune_conv2d_cuda.html | 340 +---
docs/how_to/work_with_microtvm/micro_autotune.html | 16 +-
docs/how_to/work_with_microtvm/micro_pytorch.html | 5 +-
docs/how_to/work_with_microtvm/micro_train.html | 16 +-
.../work_with_microtvm/sg_execution_times.html | 12 +-
.../how_to/work_with_relay/sg_execution_times.html | 8 +-
docs/how_to/work_with_schedules/intrin_math.html | 2 +-
.../work_with_schedules/sg_execution_times.html | 14 +-
docs/how_to/work_with_schedules/tensorize.html | 2 +-
docs/reference/api/python/auto_scheduler.html | 4 +-
.../api/typedoc/classes/bytestreamreader.html | 12 +-
.../api/typedoc/classes/cachedcallstack.html | 34 +-
docs/reference/api/typedoc/classes/dldatatype.html | 12 +-
docs/reference/api/typedoc/classes/dldevice.html | 10 +-
.../reference/api/typedoc/classes/environment.html | 12 +-
docs/reference/api/typedoc/classes/ffilibrary.html | 20 +-
.../api/typedoc/classes/graphexecutor.html | 16 +-
docs/reference/api/typedoc/classes/instance.html | 40 +-
docs/reference/api/typedoc/classes/memory.html | 34 +-
docs/reference/api/typedoc/classes/module.html | 10 +-
docs/reference/api/typedoc/classes/ndarray.html | 22 +-
.../api/typedoc/classes/packedfunccell.html | 6 +-
docs/reference/api/typedoc/classes/rpcserver.html | 14 +-
docs/reference/api/typedoc/classes/scalar.html | 6 +-
.../api/typedoc/classes/webgpucontext.html | 12 +-
docs/reference/api/typedoc/enums/argtypecode.html | 30 +-
.../api/typedoc/enums/aynccallbackcode.html | 4 +-
.../api/typedoc/enums/dldatatypecode.html | 8 +-
.../api/typedoc/enums/rpcserverstate.html | 12 +-
docs/reference/api/typedoc/enums/sizeof.html | 18 +-
docs/reference/api/typedoc/index.html | 112 +-
.../api/typedoc/interfaces/disposable.html | 2 +-
.../api/typedoc/interfaces/functioninfo.html | 6 +-
.../api/typedoc/interfaces/libraryprovider.html | 4 +-
docs/searchindex.js | 2 +-
.../vta/tutorials/autotvm/sg_execution_times.html | 6 +-
.../tutorials/frontend/deploy_classification.html | 2 +-
.../vta/tutorials/frontend/deploy_detection.html | 2 +-
.../vta/tutorials/frontend/sg_execution_times.html | 6 +-
.../vta/tutorials/optimize/sg_execution_times.html | 6 +-
docs/topic/vta/tutorials/sg_execution_times.html | 4 +-
docs/tutorial/auto_scheduler_matmul_x86.html | 4 +-
docs/tutorial/autotvm_matmul_x86.html | 20 +-
docs/tutorial/autotvm_relay_x86.html | 269 ++-
docs/tutorial/cross_compilation_and_rpc.html | 2 +-
docs/tutorial/intro_topi.html | 2 +-
docs/tutorial/sg_execution_times.html | 24 +-
docs/tutorial/tensor_expr_get_started.html | 42 +-
129 files changed, 4645 insertions(+), 2270 deletions(-)
diff --git a/docs/_images/sphx_glr_micro_train_001.png b/docs/_images/sphx_glr_micro_train_001.png
index b0d90a1aba..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 029d81d2c2..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 53ad5d9fe6..308bf8ca33 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 19.680 seconds)
+ **Total running time of the script:** ( 1 minutes 19.010 seconds)
.. _sphx_glr_download_how_to_compile_models_from_darknet.py:
diff --git a/docs/_sources/how_to/compile_models/from_keras.rst.txt b/docs/_sources/how_to/compile_models/from_keras.rst.txt
index 2cc7bdddf3..2f976ad2eb 100644
--- a/docs/_sources/how_to/compile_models/from_keras.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_keras.rst.txt
@@ -232,7 +232,7 @@ Look up prediction top 1 index in 1000 class synset.
.. code-block:: none
Relay top-1 id: 285, class name: Egyptian cat
-
1/1 [==============================] - ETA: 0s
1/1 [==============================] - 1s 928ms/step
+
1/1 [==============================] - ETA: 0s
1/1 [==============================] - 1s 937ms/step
Keras top-1 id: 285, class name: Egyptian cat
diff --git a/docs/_sources/how_to/compile_models/from_mxnet.rst.txt b/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
index 225578a2a7..455c161eac 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.zip02131915-cf90-43bd-a4e4-e8fdd67e5c71 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+ Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip533579d2-278b-4bde-84d7-3b4a17fc931c 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 32ab1e19c6..2f56bd5bc3 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 | 7.99M/41.5M [00:00<00:00, 39.4MB/s]
35%|###4 | 14.3M/41.5M [00:00<00:00, 43.6MB/s]
47%|####6 | 19.3M/41.5M [00:00<00:00, 46.5MB/s]
58%|#####7 | 23.9M/41.5M [00:00<00:00, 45.9MB/s]
68%|######8 | 28.4M/41.5M [00:00<00:00, 45.5MB/s]
79%|#######8 | 32.8M/41.5M [00:00<00:00, 38.9MB/s]
92%|#########2| 38.3M/41.5M [00:00<00:00, 40.3MB/s]
100%|##########| 41.5M/41.5M [00:01<00:00, 42.0MB/s]
+
0%| | 0.00/41.5M [00:00<?, ?B/s]
19%|#9 | 7.99M/41.5M [00:00<00:00, 49.0MB/s]
35%|###4 | 14.3M/41.5M [00:00<00:00, 39.8MB/s]
44%|####3 | 18.2M/41.5M [00:00<00:00, 38.3MB/s]
54%|#####3 | 22.3M/41.5M [00:00<00:00, 39.1MB/s]
63%|######2 | 26.1M/41.5M [00:00<00:00, 36.9MB/s]
77%|#######7 | 32.0M/41.5M [00:00<00:00, 42.7MB/s]
92%|#########2| 38.3M/41.5M [00:01<00:00, 30.5MB/s]
100%|##########| 41.5M/41.5M [00:01<00:00, 34.6MB/s]
diff --git a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
index 1f9ebe6000..c6528ae644 100644
--- a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
@@ -101,7 +101,7 @@ Load a pretrained PyTorch model
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=ResNet18_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet18_Weights.DEFAULT` to get the most up-to-date weights.
warnings.warn(msg)
Downloading: "https://download.pytorch.org/models/resnet18-f37072fd.pth" to /workspace/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
-
0%| | 0.00/44.7M [00:00<?, ?B/s]
18%|#7 | 7.99M/44.7M [00:00<00:00, 49.1MB/s]
36%|###5 | 16.0M/44.7M [00:00<00:00, 60.1MB/s]
54%|#####4 | 24.2M/44.7M [00:00<00:00, 69.7MB/s]
70%|######9 | 31.1M/44.7M [00:00<00:00, 51.5MB/s]
82%|########2 | 36.6M/44.7M [00:00<00:00, 46.7MB/s]
93%|#########2| 41.5M/44.7M [00:00<00:00, 43.4MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 49.1MB/s]
+
0%| | 0.00/44.7M [00:00<?, ?B/s]
14%|#4 | 6.30M/44.7M [00:00<00:00, 65.5MB/s]
34%|###3 | 15.2M/44.7M [00:00<00:00, 81.7MB/s]
54%|#####3 | 24.0M/44.7M [00:00<00:00, 78.4MB/s]
72%|#######1 | 32.1M/44.7M [00:00<00:00, 79.4MB/s]
90%|########9 | 40.0M/44.7M [00:00<00:00, 65.1MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 60.6MB/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 d07eada0f0..1508770873 100644
--- a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
@@ -424,7 +424,7 @@ Run the corresponding model on tensorflow
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 23.311 seconds)
+ **Total running time of the script:** ( 1 minutes 22.434 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 bb8f5b3f67..55656b934d 100644
--- a/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
@@ -5,26 +5,26 @@
Computation times
=================
-**06:34.450** total execution time for **how_to_compile_models** files:
+**06:31.248** total execution time for **how_to_compile_models** files:
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:23.311 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:22.434 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:19.680 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:19.010 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:53.899 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:53.764 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:36.783 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:36.826 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:31.246 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:31.177 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:30.816 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:30.700 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:28.290 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:27.622 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:25.268 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:25.010 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:22.520 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:22.047 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.637 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.657 | 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 ad006def0e..df3fc2a1de 100644
--- a/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt
@@ -727,7 +727,7 @@ well as provides information about the model's performance
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 2753.6120 2753.0844 2757.5755 2751.7726 1.6565
+ 2753.7866 2752.9268 2758.3577 2750.6551 2.1445
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 bd3ad8b394..69c2575f88 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)
- 15.7103 15.5860 16.5707 15.4881 0.3229
+ 16.4693 16.3055 17.3688 15.9610 0.4562
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 85d72b2e66..6417bd0253 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<00:06, 24.5MB/s]
5%|5 | 8.65M/170M [00:00<00:06, 24.5MB/s]
8%|8 | 14.3M/170M [00:00<00:05, 29.0MB/s]
10%|# | 17.1M/170M [00:00<00:06, 26.0MB/s]
13%|#3 | 22.3M/170M [00:00<00:04, 31.6MB/s]
15%|#4 | 25.4M/170M [00:00<00:05, 27.6MB/s]
19%|#8 | 32.0M/170M [00:01<00:04, 32.6MB/s]
24%|##3 | 40.0M/170M [00:01<00:03, 36.5MB/s]
28%|##8 | 48.0M/170M [00:01<00:03, 42.1MB/s]
31%|### | 52.0M/170M [00:01<00:03, 36.8MB/s]
33%|###2 | 56.0M/170M [00:01<00:03, 36.6MB/s]
37%|###6 | 62.3M/170M [00:01<00:03, 35.3MB/s]
39%|###8 | 65.7M/170M [00:02<00:03, 33.5MB/s]
42%|####2 | 72.0M/170M [00:02<00:02, 38.7MB/s]
47%|####7 | 80.0M/170M [00:02<00:01, 47.2MB/s]
52%|#####1 | 88.0M/170M [00:02<00:01, 51.7MB/s]
56%|#####5 | 94.3M/170M [00:02<00:01, 52.0MB/
s]
60%|###### | 102M/170M [00:02<00:01, 59.5MB/s]
65%|######4 | 110M/170M [00:02<00:00, 64.7MB/s]
68%|######8 | 116M/170M [00:02<00:00, 57.6MB/s]
72%|#######1 | 122M/170M [00:03<00:00, 55.3MB/s]
75%|#######5 | 128M/170M [00:03<00:00, 52.4MB/s]
80%|######## | 136M/170M [00:03<00:00, 50.2MB/s]
83%|########3 | 141M/170M [00:03<00:00, 50.5MB/s]
86%|########6 | 146M/170M [00:03<00:00, 42.7MB/s]
93%|#########3| 158M/170M [00:03<00:00, 61.5MB/s]
97%|#########7| 165M/170M [00:03<00:00, 54.6MB/s]
100%|##########| 170M/170M [00:04<00:00, 43.1MB/s]
+
0%| | 0.00/170M [00:00<?, ?B/s]
4%|3 | 6.30M/170M [00:00<00:04, 40.9MB/s]
6%|6 | 10.2M/170M [00:00<00:04, 35.6MB/s]
9%|9 | 16.0M/170M [00:00<00:04, 34.0MB/s]
13%|#3 | 22.3M/170M [00:00<00:03, 42.2MB/s]
18%|#7 | 30.3M/170M [00:00<00:02, 49.5MB/s]
21%|## | 35.2M/170M [00:00<00:03, 43.5MB/s]
27%|##7 | 46.3M/170M [00:00<00:02, 61.4MB/s]
31%|###1 | 52.7M/170M [00:01<00:02, 49.2MB/s]
34%|###4 | 58.1M/170M [00:01<00:02, 48.3MB/s]
38%|###7 | 64.0M/170M [00:01<00:02, 48.6MB/s]
42%|####2 | 72.0M/170M [00:01<00:01, 51.6MB/s]
47%|####7 | 80.0M/170M [00:01<00:01, 53.4MB/s]
52%|#####1 | 88.0M/170M [00:01<00:01, 57.0MB/s]
56%|#####5 | 94.3M/170M [00:02<00:01, 52.5MB/s]
60%|###### | 102M/170M [00:02<00:01, 58.8MB/s]
64%|######3 | 108M/170M [00:02<00:01, 56.6MB/s]
68%|######8 | 116M/170M [00:02<00:00, 63.7MB/s]
72%|#######2 | 123M/170M [00:02<00:00, 56.1MB/s]
75%|#######5 | 128M/170M [00:02<00:00, 54.7MB/s]
80%|######## | 136M/170M [00:02<00:00, 62.0MB/s]
85%|########4 | 144M/170M [00:02<00:00, 53.5MB/s]
88%|########8 | 150M/170M [00:03<00:00, 51.4MB/s]
92%|#########1| 155M/170M [00:03<00:00, 48.0MB/s]
97%|#########7| 165M/170M [00:03<00:00, 54.1MB/s]
100%|##########| 170M/170M [00:03<00:00, 50.9MB/s]
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/nn/functional.py:3897: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
for i in range(dim)
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/detection/anchor_utils.py:124: UserWarning: __floordiv__ is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor').
@@ -299,7 +299,7 @@ Get boxes with score larger than 0.9
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 3 minutes 28.687 seconds)
+ **Total running time of the script:** ( 3 minutes 27.823 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 ed8dce144e..11005bc332 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, 57.9MB/s]
100%|#########9| 13.5M/13.6M [00:00<00:00, 45.4MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 47.1MB/s]
+
0%| | 0.00/13.6M [00:00<?, ?B/s]
59%|#####8 | 7.99M/13.6M [00:00<00:00, 59.9MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 63.9MB/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)
- 90.3866 90.3266 92.8951 90.0252 0.4161
+ 90.6068 90.5499 91.6871 90.2145 0.2587
@@ -458,7 +458,7 @@ TODO
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 14.058 seconds)
+ **Total running time of the script:** ( 1 minutes 13.472 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 d1605a1ce7..9d76c6d7d2 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)
- 120.4425 120.4263 124.4521 119.7056 0.4887
+ 119.3548 119.3192 123.5348 117.3429 0.7662
@@ -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 41.952 seconds)
+ **Total running time of the script:** ( 2 minutes 32.431 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 9919320cf5..52340e4a69 100644
--- a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
@@ -257,7 +257,7 @@ We create a Relay VM to build and execute the model.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 50.624 seconds)
+ **Total running time of the script:** ( 1 minutes 50.647 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 048cd4f224..9dcd543bf6 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 | 3379/132723 [00:00<00:03, 33787.94KB/s]
8%|7 | 10298/132723 [00:00<00:02, 54608.05KB/s]
14%|#3 | 18379/132723 [00:00<00:01, 66569.98KB/s]
20%|## | 26553/132723 [00:00<00:01, 72555.86KB/s]
26%|##6 | 34718/132723 [00:00<00:01, 75824.82KB/s]
32%|###2 | 42862/132723 [00:00<00:01, 77732.51KB/s]
38%|###8 | 50928/132723 [00:00<00:01, 78686.63KB/s]
45%|####4 | 59078/132723 [00:00<00:00, 79580.64KB/s]
51%|##### | 67208/132723 [00:00<00:00, 80116.39KB/s]
57%|#####6 | 75394/132723 [00:01<00:00, 80652.35KB/s]
63%|######2 | 83533/132723 [00:01<00:00, 80876.29KB/s]
69%|######9 | 91640/132723 [00:01<00:00, 80932.71KB/s]
75%|#######5 | 99762/132723 [00:01<00:00, 81017.01KB/s]
81%|########1 | 107929/132723 [00:01<00:00, 81210.81KB/s]
87%|########7 | 116075/132723 [00:01<00:00, 81282.84KB/s]
94%|#########
3| 124220/132723 [00:01<00:00, 81331.19KB/s]
100%|#########9| 132534/132723 [00:01<00:00, 81873.09KB/s]
100%|##########| 132723/132723 [00:01<00:00, 77858.45KB/s]
+
0%| | 0/132723 [00:00<?, ?KB/s]
2%|1 | 2605/132723 [00:00<00:04, 26045.86KB/s]
6%|6 | 8383/132723 [00:00<00:02, 44657.14KB/s]
12%|#2 | 16190/132723 [00:00<00:01, 59901.02KB/s]
19%|#8 | 24805/132723 [00:00<00:01, 70256.82KB/s]
25%|##4 | 32879/132723 [00:00<00:01, 74030.58KB/s]
31%|###1 | 41471/132723 [00:00<00:01, 78068.40KB/s]
38%|###7 | 50022/132723 [00:00<00:01, 80496.63KB/s]
44%|####4 | 58660/132723 [00:00<00:00, 82367.69KB/s]
51%|##### | 67243/132723 [00:00<00:00, 83446.90KB/s]
57%|#####7 | 75896/132723 [00:01<00:00, 84397.08KB/s]
64%|######3 | 84525/132723 [00:01<00:00, 84973.34KB/s]
70%|####### | 93215/132723 [00:01<00:00, 85555.58KB/s]
77%|#######6 | 101849/132723 [00:01<00:00, 85790.86KB/s]
83%|########3 | 110556/132723 [00:01<00:00, 86171.44KB/s]
90%|########9 | 119208/132723 [00:01<00:00, 86274.60KB/s]
96%|#########
6| 127911/132723 [00:01<00:00, 86500.86KB/s]
100%|##########| 132723/132723 [00:01<00:00, 80060.59KB/s]
@@ -246,7 +246,7 @@ Display result
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 3 minutes 43.106 seconds)
+ **Total running time of the script:** ( 3 minutes 42.134 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 6ff2df69ff..c0603cf501 100644
--- a/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt
@@ -5,26 +5,26 @@
Computation times
=================
-**15:29.939** total execution time for **how_to_deploy_models** files:
+**15:17.691** total execution time for **how_to_deploy_models** files:
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 03:43.106 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 03:42.134 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:28.687 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:27.823 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 02:41.952 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 02:32.431 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:50.624 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:50.647 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:14.058 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:13.472 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``) | 00:56.124 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``) | 00:56.069 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:40.685 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:40.441 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:27.490 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:27.542 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:27.207 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:27.127 | 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 795f5acb19..8dba772b26 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.zip29b949b8-121d-4be9-864b-50e9878edb73 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+ Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zipc9c140f0-8012-4216-bf1a-486ce56b5235 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 c478faa74b..afff367c78 100644
--- a/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt
@@ -5,14 +5,14 @@
Computation times
=================
-**00:53.009** total execution time for **how_to_extend_tvm** files:
+**00:52.742** total execution time for **how_to_extend_tvm** files:
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:49.188 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:48.972 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.733 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.697 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:01.079 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:01.064 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_extend_tvm_low_level_custom_pass.py` (``low_level_custom_pass.py``) | 00:00.008 | 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 4acfc780cc..372ed73412 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: 22161us [22161us] (48.95%; 48.95%)
- FoldScaleAxis: 23109us [7us] (51.05%; 51.05%)
- FoldConstant: 23102us [1705us] (51.03%; 99.97%)
- InferType: 21397us [21397us] (47.26%; 92.62%)
+ InferType: 22397us [22397us] (48.70%; 48.70%)
+ FoldScaleAxis: 23591us [7us] (51.30%; 51.30%)
+ FoldConstant: 23583us [1737us] (51.28%; 99.97%)
+ InferType: 21847us [21847us] (47.50%; 92.64%)
@@ -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: 21509us [21509us] (48.47%; 48.47%)
- FoldScaleAxis: 22864us [5us] (51.53%; 51.53%)
- FoldConstant: 22859us [1659us] (51.52%; 99.98%)
- InferType: 21200us [21200us] (47.78%; 92.74%)
+ InferType: 21925us [21925us] (48.29%; 48.29%)
+ FoldScaleAxis: 23478us [6us] (51.71%; 51.71%)
+ FoldConstant: 23473us [1741us] (51.70%; 99.98%)
+ InferType: 21732us [21732us] (47.86%; 92.58%)
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 320c4ab464..b877d3a5ba 100644
--- a/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
@@ -331,7 +331,7 @@ latency of convolution.
.. code-block:: none
- Convolution: 54.091423 ms
+ Convolution: 33.637153 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 5e2814f320..f8fa5525f0 100644
--- a/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt
@@ -608,7 +608,7 @@ be able to run on our build server
.. code-block:: none
- conv2d with tensor core: 6.593321 ms
+ conv2d with tensor core: 13.335760 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 de9e0fc716..e1d297dbdf 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: 0.018411
- Baseline: 3.234937
+ Numpy running time: 0.018666
+ Baseline: 3.252249
@@ -227,7 +227,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
.. code-block:: none
- Opt1: 0.299118
+ Opt1: 0.301624
@@ -318,7 +318,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
.. code-block:: none
- Opt2: 0.333509
+ Opt2: 0.335315
@@ -406,7 +406,7 @@ the access pattern for A matrix is more cache friendly.
.. code-block:: none
- Opt3: 0.115660
+ Opt3: 0.114184
@@ -523,7 +523,7 @@ flattening.
.. code-block:: none
- Opt4: 0.109347
+ Opt4: 0.109062
@@ -635,7 +635,7 @@ write to C when all the block results are ready.
.. code-block:: none
- Opt5: 0.110709
+ Opt5: 0.114539
@@ -748,7 +748,7 @@ Furthermore, we can also utilize multi-core processors to do the thread-level pa
.. code-block:: none
- Opt6: 0.146349
+ Opt6: 0.151849
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 4feb9a76d2..ed561cacf7 100644
--- a/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
@@ -5,12 +5,12 @@
Computation times
=================
-**00:34.251** total execution time for **how_to_optimize_operators** files:
+**00:34.787** total execution time for **how_to_optimize_operators** files:
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:31.679 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:32.109 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.485 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.598 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.087 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.080 | 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 23c9315597..ff3143c6ea 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt
@@ -5,18 +5,18 @@
Computation times
=================
-**09:47.420** total execution time for **how_to_tune_with_autoscheduler** files:
+**09:54.297** total execution time for **how_to_tune_with_autoscheduler** files:
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 05:59.876 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 06:08.809 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:40.136 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:40.129 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 01:07.619 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 01:07.165 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:31.899 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:30.521 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:14.637 | 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_mali.py` (``tune_network_mali.py``) | 00:13.252 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:13.099 | 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 81c32b0c1f..07d6598fe9 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt
@@ -244,181 +244,797 @@ cooperative fetching, unrolling and operator fusion.
def main(data: T.Buffer((1, 512, 7, 7), "float32"), kernel: T.Buffer((512, 512, 3, 3), "float32"), bias: T.Buffer((1, 512, 1, 1), "float32"), compute: T.Buffer((1, 512, 7, 7), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
blockIdx_x = T.env_thread("blockIdx.x")
- T.launch_thread(blockIdx_x, 16)
- conv2d_nchw = T.allocate([2], "float32", "local")
- pad_temp_shared = T.allocate([648], "float32", "shared")
+ T.launch_thread(blockIdx_x, 56)
+ conv2d_nchw = T.allocate([16], "float32", "local")
+ pad_temp_shared = T.allocate([108], "float32", "shared")
kernel_shared = T.allocate([2304], "float32", "shared")
threadIdx_x = T.env_thread("threadIdx.x")
- T.launch_thread(threadIdx_x, 784)
- conv2d_nchw_1 = T.Buffer((2,), data=conv2d_nchw, scope="local", align=8)
+ T.launch_thread(threadIdx_x, 28)
+ conv2d_nchw_1 = T.Buffer((16,), data=conv2d_nchw, scope="local", align=16)
conv2d_nchw_1[0] = T.float32(0)
+ conv2d_nchw_1[4] = T.float32(0)
+ conv2d_nchw_1[8] = T.float32(0)
+ conv2d_nchw_1[12] = T.float32(0)
conv2d_nchw_1[1] = T.float32(0)
- for rc_outer_outer in range(64):
- cse_var_1: T.int32 = rc_outer_outer * 72
+ conv2d_nchw_1[5] = T.float32(0)
+ conv2d_nchw_1[9] = T.float32(0)
+ conv2d_nchw_1[13] = T.float32(0)
+ conv2d_nchw_1[2] = T.float32(0)
+ conv2d_nchw_1[6] = T.float32(0)
+ conv2d_nchw_1[10] = T.float32(0)
+ conv2d_nchw_1[14] = T.float32(0)
+ conv2d_nchw_1[3] = T.float32(0)
+ conv2d_nchw_1[7] = T.float32(0)
+ conv2d_nchw_1[11] = T.float32(0)
+ conv2d_nchw_1[15] = T.float32(0)
+ for rc_outer_outer in range(128):
+ cse_var_2: T.int32 = rc_outer_outer * 196
+ cse_var_1: T.int32 = rc_outer_outer * 36
threadIdx_x_1 = T.env_thread("threadIdx.x")
- pad_temp_shared_1 = T.Buffer((648,), data=pad_temp_shared, scope="shared")
- with T.launch_thread(threadIdx_x_1, 784):
- if T.likely(threadIdx_x_1 < 648):
- data_1 = T.Buffer((25088,), data=data.data)
- pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(9 <= threadIdx_x_1 % 81 and threadIdx_x_1 % 81 < 72 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[rc_outer_outer * 392 + threadIdx_x_1 // 81 * 49 + threadIdx_x_1 % 81 // 9 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
+ pad_temp_shared_1 = T.Buffer((108,), data=pad_temp_shared, scope="shared")
+ data_1 = T.Buffer((25088,), data=data.data)
+ with T.launch_thread(threadIdx_x_1, 28):
+ pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(1 <= threadIdx_x_1 % 27 // 9 + blockIdx_x % 7 and threadIdx_x_1 % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + threadIdx_x_1 // 27 * 49 + threadIdx_x_1 % 27 // 9 * 7 + blockIdx_x % 7 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 28):
+ pad_temp_shared_1[threadIdx_x_1 + 28] = T.if_then_else(1 <= (threadIdx_x_1 + 1) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 1) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 1) % 9 and (threadIdx_x_1 + 1) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 28) // 27 * 49 + (threadIdx_x_1 + 1) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 1) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 28):
+ pad_temp_shared_1[threadIdx_x_1 + 56] = T.if_then_else(1 <= (threadIdx_x_1 + 2) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 2) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 2) % 9 and (threadIdx_x_1 + 2) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 56) // 27 * 49 + (threadIdx_x_1 + 2) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 2) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 28):
+ if T.likely(threadIdx_x_1 < 24):
+ pad_temp_shared_1[threadIdx_x_1 + 84] = T.if_then_else(1 <= (threadIdx_x_1 + 3) // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 3) // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 3) % 9 and (threadIdx_x_1 + 3) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 84) // 27 * 49 + (threadIdx_x_1 + 3) // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 3) % 9 - 8], T.float32(0))
threadIdx_x_2 = T.env_thread("threadIdx.x")
kernel_shared_1 = T.Buffer((2304,), data=kernel_shared, scope="shared")
kernel_1 = T.Buffer((2359296,), data=kernel.data)
- with T.launch_thread(threadIdx_x_2, 784):
- kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x * 147456 + threadIdx_x_2 // 72 * 4608 + cse_var_1 + threadIdx_x_2 % 72]
- with T.launch_thread(threadIdx_x_2, 784):
- kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x * 147456 + (threadIdx_x_2 + 784) // 72 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 72 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 784):
- if T.likely(threadIdx_x_2 < 736):
- kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x * 147456 + (threadIdx_x_2 + 1568) // 72 * 4608 + cse_var_1 + (threadIdx_x_2 + 56) % 72 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 49 * 144]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 49 * 144 + 72]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 49 * 144 + 1]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 49 * 144 + 73]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 49 * 144 + 2]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 49 * 144 + 74]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 49 * 144 + 9]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 49 * 144 + 81]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 49 * 144 + 10]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 49 * 144 + 82]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 49 * 144 + 11]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 49 * 144 + 83]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 162] * kernel_shared_1[threadIdx_x // 49 * 144 + 18]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 162] * kernel_shared_1[threadIdx_x // 49 * 144 + 90]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 163] * kernel_shared_1[threadIdx_x // 49 * 144 + 19]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 163] * kernel_shared_1[threadIdx_x // 49 * 144 + 91]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 164] * kernel_shared_1[threadIdx_x // 49 * 144 + 20]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 164] * kernel_shared_1[threadIdx_x // 49 * 144 + 92]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 243] * kernel_shared_1[threadIdx_x // 49 * 144 + 27]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 243] * kernel_shared_1[threadIdx_x // 49 * 144 + 99]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 244] * kernel_shared_1[threadIdx_x // 49 * 144 + 28]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 244] * kernel_shared_1[threadIdx_x // 49 * 144 + 100]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 245] * kernel_shared_1[threadIdx_x // 49 * 144 + 29]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 245] * kernel_shared_1[threadIdx_x // 49 * 144 + 101]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 324] * kernel_shared_1[threadIdx_x // 49 * 144 + 36]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 324] * kernel_shared_1[threadIdx_x // 49 * 144 + 108]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 325] * kernel_shared_1[threadIdx_x // 49 * 144 + 37]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 325] * kernel_shared_1[threadIdx_x // 49 * 144 + 109]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 326] * kernel_shared_1[threadIdx_x // 49 * 144 + 38]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 326] * kernel_shared_1[threadIdx_x // 49 * 144 + 110]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 405] * kernel_shared_1[threadIdx_x // 49 * 144 + 45]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 405] * kernel_shared_1[threadIdx_x // 49 * 144 + 117]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 406] * kernel_shared_1[threadIdx_x // 49 * 144 + 46]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 406] * kernel_shared_1[threadIdx_x // 49 * 144 + 118]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 407] * kernel_shared_1[threadIdx_x // 49 * 144 + 47]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 407] * kernel_shared_1[threadIdx_x // 49 * 144 + 119]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 486] * kernel_shared_1[threadIdx_x // 49 * 144 + 54]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 486] * kernel_shared_1[threadIdx_x // 49 * 144 + 126]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 487] * kernel_shared_1[threadIdx_x // 49 * 144 + 55]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 487] * kernel_shared_1[threadIdx_x // 49 * 144 + 127]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 488] * kernel_shared_1[threadIdx_x // 49 * 144 + 56]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 488] * kernel_shared_1[threadIdx_x // 49 * 144 + 128]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 567] * kernel_shared_1[threadIdx_x // 49 * 144 + 63]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 567] * kernel_shared_1[threadIdx_x // 49 * 144 + 135]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 568] * kernel_shared_1[threadIdx_x // 49 * 144 + 64]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 568] * kernel_shared_1[threadIdx_x // 49 * 144 + 136]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 569] * kernel_shared_1[threadIdx_x // 49 * 144 + 65]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 569] * kernel_shared_1[threadIdx_x // 49 * 144 + 137]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 49 * 144 + 3]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 49 * 144 + 75]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 49 * 144 + 4]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 49 * 144 + 76]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 49 * 144 + 5]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 49 * 144 + 77]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 49 * 144 + 12]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 49 * 144 + 84]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 49 * 144 + 13]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 49 * 144 + 85]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 49 * 144 + 14]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 49 * 144 + 86]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 171] * kernel_shared_1[threadIdx_x // 49 * 144 + 21]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 171] * kernel_shared_1[threadIdx_x // 49 * 144 + 93]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 172] * kernel_shared_1[threadIdx_x // 49 * 144 + 22]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 172] * kernel_shared_1[threadIdx_x // 49 * 144 + 94]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 173] * kernel_shared_1[threadIdx_x // 49 * 144 + 23]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 173] * kernel_shared_1[threadIdx_x // 49 * 144 + 95]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 252] * kernel_shared_1[threadIdx_x // 49 * 144 + 30]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 252] * kernel_shared_1[threadIdx_x // 49 * 144 + 102]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 253] * kernel_shared_1[threadIdx_x // 49 * 144 + 31]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 253] * kernel_shared_1[threadIdx_x // 49 * 144 + 103]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 254] * kernel_shared_1[threadIdx_x // 49 * 144 + 32]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 254] * kernel_shared_1[threadIdx_x // 49 * 144 + 104]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 333] * kernel_shared_1[threadIdx_x // 49 * 144 + 39]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 333] * kernel_shared_1[threadIdx_x // 49 * 144 + 111]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 334] * kernel_shared_1[threadIdx_x // 49 * 144 + 40]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 334] * kernel_shared_1[threadIdx_x // 49 * 144 + 112]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 335] * kernel_shared_1[threadIdx_x // 49 * 144 + 41]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 335] * kernel_shared_1[threadIdx_x // 49 * 144 + 113]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 414] * kernel_shared_1[threadIdx_x // 49 * 144 + 48]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 414] * kernel_shared_1[threadIdx_x // 49 * 144 + 120]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 415] * kernel_shared_1[threadIdx_x // 49 * 144 + 49]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 415] * kernel_shared_1[threadIdx_x // 49 * 144 + 121]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 416] * kernel_shared_1[threadIdx_x // 49 * 144 + 50]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 416] * kernel_shared_1[threadIdx_x // 49 * 144 + 122]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 495] * kernel_shared_1[threadIdx_x // 49 * 144 + 57]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 495] * kernel_shared_1[threadIdx_x // 49 * 144 + 129]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 496] * kernel_shared_1[threadIdx_x // 49 * 144 + 58]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 496] * kernel_shared_1[threadIdx_x // 49 * 144 + 130]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 497] * kernel_shared_1[threadIdx_x // 49 * 144 + 59]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 497] * kernel_shared_1[threadIdx_x // 49 * 144 + 131]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 576] * kernel_shared_1[threadIdx_x // 49 * 144 + 66]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 576] * kernel_shared_1[threadIdx_x // 49 * 144 + 138]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 577] * kernel_shared_1[threadIdx_x // 49 * 144 + 67]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 577] * kernel_shared_1[threadIdx_x // 49 * 144 + 139]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 578] * kernel_shared_1[threadIdx_x // 49 * 144 + 68]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 578] * kernel_shared_1[threadIdx_x // 49 * 144 + 140]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 49 * 144 + 6]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 49 * 144 + 78]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 49 * 144 + 7]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 49 * 144 + 79]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 49 * 144 + 8]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 49 * 144 + 80]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 49 * 144 + 15]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 49 * 144 + 87]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 49 * 144 + 16]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 49 * 144 + 88]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 49 * 144 + 17]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 49 * 144 + 89]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 180] * kernel_shared_1[threadIdx_x // 49 * 144 + 24]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 180] * kernel_shared_1[threadIdx_x // 49 * 144 + 96]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 181] * kernel_shared_1[threadIdx_x // 49 * 144 + 25]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 181] * kernel_shared_1[threadIdx_x // 49 * 144 + 97]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 182] * kernel_shared_1[threadIdx_x // 49 * 144 + 26]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 182] * kernel_shared_1[threadIdx_x // 49 * 144 + 98]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 261] * kernel_shared_1[threadIdx_x // 49 * 144 + 33]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 261] * kernel_shared_1[threadIdx_x // 49 * 144 + 105]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 262] * kernel_shared_1[threadIdx_x // 49 * 144 + 34]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 262] * kernel_shared_1[threadIdx_x // 49 * 144 + 106]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 263] * kernel_shared_1[threadIdx_x // 49 * 144 + 35]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 263] * kernel_shared_1[threadIdx_x // 49 * 144 + 107]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 342] * kernel_shared_1[threadIdx_x // 49 * 144 + 42]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 342] * kernel_shared_1[threadIdx_x // 49 * 144 + 114]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 343] * kernel_shared_1[threadIdx_x // 49 * 144 + 43]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 343] * kernel_shared_1[threadIdx_x // 49 * 144 + 115]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 344] * kernel_shared_1[threadIdx_x // 49 * 144 + 44]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 344] * kernel_shared_1[threadIdx_x // 49 * 144 + 116]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 423] * kernel_shared_1[threadIdx_x // 49 * 144 + 51]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 423] * kernel_shared_1[threadIdx_x // 49 * 144 + 123]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 424] * kernel_shared_1[threadIdx_x // 49 * 144 + 52]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 424] * kernel_shared_1[threadIdx_x // 49 * 144 + 124]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 425] * kernel_shared_1[threadIdx_x // 49 * 144 + 53]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 425] * kernel_shared_1[threadIdx_x // 49 * 144 + 125]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 504] * kernel_shared_1[threadIdx_x // 49 * 144 + 60]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 504] * kernel_shared_1[threadIdx_x // 49 * 144 + 132]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 505] * kernel_shared_1[threadIdx_x // 49 * 144 + 61]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 505] * kernel_shared_1[threadIdx_x // 49 * 144 + 133]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 506] * kernel_shared_1[threadIdx_x // 49 * 144 + 62]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 506] * kernel_shared_1[threadIdx_x // 49 * 144 + 134]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 585] * kernel_shared_1[threadIdx_x // 49 * 144 + 69]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 585] * kernel_shared_1[threadIdx_x // 49 * 144 + 141]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 586] * kernel_shared_1[threadIdx_x // 49 * 144 + 70]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 586] * kernel_shared_1[threadIdx_x // 49 * 144 + 142]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 587] * kernel_shared_1[threadIdx_x // 49 * 144 + 71]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 587] * kernel_shared_1[threadIdx_x // 49 * 144 + 143]
- for i1_inner in range(2):
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 28] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 28) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 56] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 56) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 84] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 84) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 112] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 112) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 140] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 140) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 168] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 168) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 196] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 196) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 224) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 252] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 32256]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 280] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 280) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 308] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 308) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 336] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 336) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 364] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 364) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 392] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 392) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 420] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 420) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 448) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 476] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 476) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 504] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 64512]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 532] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 532) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 560] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 560) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 588] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 588) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 616] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 616) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 644] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 644) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 672) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 700] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 700) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 728] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 728) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 756] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 96768]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 784) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 812] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 812) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 840] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 840) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 868] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 868) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 896) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 924] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 924) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 952] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 952) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 980] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 980) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1008] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 129024]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1036] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1036) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1064] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1064) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1092] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1092) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1120) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1148] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1148) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1176] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1176) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1204] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1204) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1232] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1232) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1260] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 161280]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1288] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1288) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1316] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1316) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1344) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1372] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1372) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1400] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1400) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1428] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1428) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1456] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1456) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1484] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1484) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1512] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 193536]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1540] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1540) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1568) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1596] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1596) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1624] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1624) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1652] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1652) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1680] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1680) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1708] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1708) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1736] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1736) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1764] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 225792]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1792) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1820] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1820) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1848] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1848) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1876] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1876) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1904] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1904) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1932] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1932) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1960] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1960) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1988] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1988) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 258048]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2044] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2044) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2072] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2072) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2100] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2100) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2128] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2128) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2156] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2156) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2184] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2184) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2212] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2212) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2240] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2240) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2268] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 290304]
+ with T.launch_thread(threadIdx_x_2, 28):
+ if T.likely(threadIdx_x_2 < 8):
+ kernel_shared_1[threadIdx_x_2 + 2296] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2296) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 28]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 576]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1152]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1728]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 36]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 612]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1188]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1764]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 577]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1153]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1729]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 37]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 613]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1189]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1765]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 2]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 578]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1154]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1730]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 38]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 614]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1190]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1766]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 3]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 579]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1155]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1731]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 39]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 615]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1191]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1767]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 4]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 580]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1156]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1732]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 40]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 616]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1192]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1768]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 5]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 581]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1157]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1733]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 41]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 617]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1193]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1769]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 6]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 582]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1158]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1734]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 42]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 618]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1194]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1770]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 7]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 583]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1159]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1735]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 43]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 619]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1195]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1771]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 8]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 584]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1160]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1736]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 44]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 620]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1196]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1772]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 72]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 648]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1224]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1800]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 108]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 684]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1260]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1836]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 73]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 649]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1225]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1801]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 109]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 685]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1261]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1837]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 74]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 650]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1226]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1802]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 110]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 686]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1262]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1838]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 75]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 651]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1227]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1803]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 111]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 687]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1263]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1839]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 76]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 652]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1228]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1804]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 112]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 688]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1264]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1840]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 77]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 653]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1229]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1805]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 113]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 689]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1265]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1841]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 78]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 654]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1230]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1806]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 114]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 690]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1266]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1842]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 79]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 655]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1231]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1807]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 115]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 691]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1267]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1843]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 80]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 656]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1232]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1808]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 116]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 692]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1268]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1844]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 9]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 585]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1161]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1737]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 45]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 621]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1197]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1773]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 10]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 586]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1162]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1738]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 46]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 622]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1198]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1774]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 11]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 587]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1163]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1739]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 47]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 623]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1199]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1775]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 12]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 588]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1164]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1740]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 48]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 624]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1200]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1776]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 13]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 589]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1165]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1741]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 49]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 625]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1201]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1777]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 14]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 590]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1166]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1742]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 50]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 626]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1202]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1778]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 15]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 591]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1167]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1743]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 51]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 627]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1203]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1779]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 16]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 592]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1168]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1744]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 52]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 628]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1204]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1780]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 17]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 593]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1169]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1745]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 53]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 629]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1205]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1781]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 81]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 657]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1233]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1809]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 117]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 693]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1269]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1845]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 82]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 658]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1234]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1810]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 118]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 694]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1270]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1846]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 83]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 659]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1235]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1811]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 119]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 695]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1271]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1847]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 84]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 660]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1236]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1812]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 120]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 696]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1272]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1848]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 85]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 661]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1237]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1813]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 121]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 697]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1273]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1849]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 86]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 662]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1238]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1814]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 122]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 698]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1274]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1850]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 87]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 663]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1239]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1815]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 123]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 699]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1275]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1851]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 88]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 664]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1240]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1816]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 124]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 700]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1276]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1852]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 89]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 665]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1241]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1817]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 125]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 701]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1277]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1853]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 18]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 594]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1170]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1746]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 54]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 630]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1206]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1782]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 19]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 595]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1171]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1747]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 55]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 631]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1207]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1783]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 20]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 596]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1172]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1748]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 56]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 632]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1208]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1784]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 21]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 597]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1173]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1749]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 57]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 633]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1209]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1785]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 22]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 598]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1174]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1750]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 58]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 634]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1210]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1786]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 23]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 599]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1175]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1751]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 59]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 635]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1211]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1787]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 24]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 600]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1176]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1752]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 60]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 636]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1212]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1788]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 25]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 601]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1177]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1753]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 61]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 637]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1213]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1789]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 26]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 602]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1178]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1754]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 62]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 638]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1214]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1790]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 90]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 666]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1242]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1818]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 126]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 702]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1278]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1854]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 91]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 667]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1243]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1819]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 127]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 703]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1279]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1855]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 92]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 668]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1244]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1820]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 128]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 704]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1280]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1856]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 93]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 669]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1245]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1821]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 129]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 705]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1281]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1857]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 94]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 670]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1246]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1822]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 130]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 706]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1282]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1858]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 95]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 671]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1247]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1823]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 131]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 707]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1283]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1859]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 96]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 672]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1248]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1824]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 132]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 708]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1284]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1860]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 97]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 673]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1249]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1825]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 133]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 709]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1285]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1861]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 98]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 674]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1250]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1826]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 134]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 710]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1286]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1862]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 27]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 603]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1179]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1755]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 63]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 639]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1215]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1791]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 28]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 604]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1180]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1756]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 64]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 640]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1216]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1792]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 29]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 605]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1181]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1757]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 65]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 641]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1217]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1793]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 30]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 606]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1182]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1758]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 66]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 642]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1218]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1794]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 31]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 607]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1183]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1759]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 67]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 643]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1219]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1795]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 32]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 608]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1184]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1760]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 68]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 644]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1220]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1796]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 33]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 609]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1185]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1761]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 69]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 645]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1221]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1797]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 34]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 610]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1186]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1762]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 70]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 646]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1222]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1798]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 35]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 611]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1187]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1763]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 71]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 647]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1223]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1799]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 99]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 675]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1251]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1827]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 135]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 711]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1287]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1863]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 100]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 676]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1252]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1828]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 136]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 712]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1288]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1864]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 101]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 677]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1253]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1829]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 137]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 713]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1289]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1865]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 102]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 678]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1254]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1830]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 138]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 714]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1290]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1866]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 103]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 679]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1255]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1831]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 139]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 715]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1291]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1867]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 104]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 680]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1256]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1832]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 140]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 716]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1292]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1868]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 105]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 681]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1257]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1833]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 141]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 717]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1293]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1869]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 106]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 682]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1258]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1834]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 142]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 718]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1294]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1870]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 107]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 683]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1259]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1835]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 143]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 719]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1295]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1871]
+ for i1_inner in range(4):
compute_1 = T.Buffer((25088,), data=compute.data)
bias_1 = T.Buffer((512,), data=bias.data)
- compute_1[blockIdx_x * 1568 + threadIdx_x // 49 * 98 + i1_inner * 49 + threadIdx_x % 49] = T.max(conv2d_nchw_1[i1_inner] + bias_1[blockIdx_x * 32 + threadIdx_x // 49 * 2 + i1_inner], T.float32(0))
+ compute_1[blockIdx_x // 7 * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + blockIdx_x % 7 * 7 + threadIdx_x % 7] = T.max(conv2d_nchw_1[i1_inner] + bias_1[blockIdx_x // 7 * 64 + threadIdx_x // 7 * 4 + i1_inner], T.float32(0))
+ compute_1[blockIdx_x // 7 * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + blockIdx_x % 7 * 7 + threadIdx_x % 7 + 784] = T.max(conv2d_nchw_1[i1_inner + 4] + bias_1[blockIdx_x // 7 * 64 + threadIdx_x // 7 * 4 + i1_inner + 16], T.float32(0))
+ compute_1[blockIdx_x // 7 * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + blockIdx_x % 7 * 7 + threadIdx_x % 7 + 1568] = T.max(conv2d_nchw_1[i1_inner + 8] + bias_1[blockIdx_x // 7 * 64 + threadIdx_x // 7 * 4 + i1_inner + 32], T.float32(0))
+ compute_1[blockIdx_x // 7 * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + blockIdx_x % 7 * 7 + threadIdx_x % 7 + 2352] = T.max(conv2d_nchw_1[i1_inner + 12] + bias_1[blockIdx_x // 7 * 64 + threadIdx_x // 7 * 4 + i1_inner + 48], T.float32(0))
@@ -468,7 +1084,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 0.383 ms
+ Execution time of this operator: 0.357 ms
@@ -517,32 +1133,32 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=2)
- conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=1)
- conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=16)
- conv2d_nchw_ff_o_o_o_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_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=2)
+ conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=4)
+ 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=4)
conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
- conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=7)
+ conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=1)
conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=1)
conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=1)
conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=7)
conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=1)
- conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=8)
- conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=1)
- conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
- conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=3)
+ conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=1)
+ conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=4)
+ conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=3)
+ conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=3)
conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=1)
s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2 [...]
compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
- compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=2)
- compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=16)
- compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+ compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=4)
+ compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=4)
+ compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=4)
compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
- compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
+ compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=1)
compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=7)
@@ -565,14 +1181,14 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[kernel_shared].fuse(kernel_shared_ax0, kernel_shared_ax1, kernel_shared_ax2, kernel_shared_ax3)
kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
s[kernel_shared].vectorize(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
- kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=784)
+ kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=28)
s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[pad_temp_shared].fuse(pad_temp_shared_ax0, pad_temp_shared_ax1, pad_temp_shared_ax2, pad_temp_shared_ax3)
pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
- pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=784)
+ 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=28)
s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
- s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
+ s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 1024)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
CUDA source code:
@@ -590,170 +1206,702 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
#define int64_t long long
#define uint64_t unsigned long long
#endif
- extern "C" __global__ void __launch_bounds__(784) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[2];
- __shared__ float pad_temp_shared[648];
+ extern "C" __global__ void __launch_bounds__(28) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
+ float conv2d_nchw[16];
+ __shared__ float pad_temp_shared[108];
__shared__ float kernel_shared[2304];
conv2d_nchw[0] = 0.000000e+00f;
+ conv2d_nchw[4] = 0.000000e+00f;
+ conv2d_nchw[8] = 0.000000e+00f;
+ conv2d_nchw[12] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
- for (int rc_outer_outer = 0; rc_outer_outer < 64; ++rc_outer_outer) {
+ conv2d_nchw[5] = 0.000000e+00f;
+ conv2d_nchw[9] = 0.000000e+00f;
+ conv2d_nchw[13] = 0.000000e+00f;
+ conv2d_nchw[2] = 0.000000e+00f;
+ conv2d_nchw[6] = 0.000000e+00f;
+ conv2d_nchw[10] = 0.000000e+00f;
+ conv2d_nchw[14] = 0.000000e+00f;
+ conv2d_nchw[3] = 0.000000e+00f;
+ conv2d_nchw[7] = 0.000000e+00f;
+ conv2d_nchw[11] = 0.000000e+00f;
+ conv2d_nchw[15] = 0.000000e+00f;
+ for (int rc_outer_outer = 0; rc_outer_outer < 128; ++rc_outer_outer) {
__syncthreads();
- if (((int)threadIdx.x) < 648) {
- pad_temp_shared[((int)threadIdx.x)] = (((((9 <= (((int)threadIdx.x) % 81)) && ((((int)threadIdx.x) % 81) < 72)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 81) * 49)) + (((((int)threadIdx.x) % 81) / 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[((int)threadIdx.x)] = (((((1 <= (((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7))) && ((((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((((rc_outer_outer * 196) + ((((int)threadIdx.x) / 27) * 49)) + (((((int)threadIdx.x) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 28)] = (((((1 <= ((((((int)threadIdx.x) + 1) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 1) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 28) / 27) * 49)) + ((((((int)threadIdx.x) + 1) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 1) % 9)) - 8)] : 0.0 [...]
+ pad_temp_shared[(((int)threadIdx.x) + 56)] = (((((1 <= ((((((int)threadIdx.x) + 2) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 2) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 56) / 27) * 49)) + ((((((int)threadIdx.x) + 2) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.0 [...]
+ if (((int)threadIdx.x) < 24) {
+ pad_temp_shared[(((int)threadIdx.x) + 84)] = (((((1 <= (((((int)threadIdx.x) + 3) / 9) + (((int)blockIdx.x) % 7))) && ((((((int)threadIdx.x) + 3) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[((((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 84) / 27) * 49)) + (((((int)threadIdx.x) + 3) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
}
- kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) * 147456) + ((((int)threadIdx.x) / 72) * 4608)) + (rc_outer_outer * 72)) + (((int)threadIdx.x) % 72))];
- kernel_shared[(((int)threadIdx.x) + 784)] = kernel[(((((((int)blockIdx.x) * 147456) + (((((int)threadIdx.x) + 784) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- if (((int)threadIdx.x) < 736) {
- kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[(((((((int)blockIdx.x) * 147456) + (((((int)threadIdx.x) + 1568) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x))];
+ kernel_shared[(((int)threadIdx.x) + 28)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 28) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 56)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 56) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 84)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 84) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 112)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 112) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 140)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 140) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 168)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 168) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 196)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 196) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 224)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 224) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 252)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 32256)];
+ kernel_shared[(((int)threadIdx.x) + 280)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 280) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 308)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 308) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 336)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 336) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 364)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 364) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 392)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 392) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 420)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 420) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 448) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 476)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 476) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 504)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 64512)];
+ kernel_shared[(((int)threadIdx.x) + 532)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 532) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 560)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 560) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 588)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 588) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 616)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 616) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 644)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 644) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 672)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 672) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 700)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 700) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 728)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 728) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 756)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 96768)];
+ kernel_shared[(((int)threadIdx.x) + 784)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 784) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 812)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 812) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 840)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 840) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 868)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 868) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 896) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 924)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 924) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 952)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 952) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 980)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 980) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 1008)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 129024)];
+ kernel_shared[(((int)threadIdx.x) + 1036)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1036) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1064)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1064) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1092)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1092) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1120) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 1148)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1148) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1176)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1176) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1204)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1204) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1232)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1232) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 1260)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 161280)];
+ kernel_shared[(((int)threadIdx.x) + 1288)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1288) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1316)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1316) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1344) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1372)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1372) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 1400)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1400) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1428)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1428) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1456)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1456) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1484)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1484) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 1512)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 193536)];
+ kernel_shared[(((int)threadIdx.x) + 1540)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1540) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1568) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1596)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1596) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1624)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1624) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 1652)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1652) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1680)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1680) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1708)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1708) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1736)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1736) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 1764)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 225792)];
+ kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1792) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1820)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1820) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1848)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1848) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1876)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1876) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 1904)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1904) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1932)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1932) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1960)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1960) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1988)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1988) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 258048)];
+ kernel_shared[(((int)threadIdx.x) + 2044)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2044) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2072)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2072) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2100)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2100) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2128)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2128) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 2156)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2156) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2184)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2184) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2212)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2212) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2240) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 2268)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 290304)];
+ if (((int)threadIdx.x) < 8) {
+ kernel_shared[(((int)threadIdx.x) + 2296)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2296) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 28)];
}
__syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[((((int)threadIdx.x) / 49) * 144)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 72)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 73)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 74)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 81)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 82)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 83)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 90)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 91)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 92)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 27)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 99)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 28)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 100)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 29)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 101)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 324)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 36)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 324)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 108)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 325)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 37)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 325)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 109)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 326)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 38)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 326)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 110)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 405)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 45)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 405)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 117)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 406)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 46)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 406)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 118)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 407)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 47)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 407)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 119)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 486)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 54)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 486)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 126)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 487)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 55)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 487)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 127)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 488)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 56)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 488)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 128)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 567)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 63)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 567)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 135)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 568)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 64)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 568)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 136)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 569)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 65)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 569)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 137)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 75)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 76)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 77)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 84)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 85)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 86)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 93)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 94)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 95)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 30)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 102)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 31)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 103)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 32)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 104)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 333)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 39)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 333)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 111)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 334)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 40)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 334)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 112)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 335)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 41)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 335)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 113)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 414)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 48)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 414)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 120)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 415)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 49)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 415)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 121)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 416)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 50)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 416)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 122)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 495)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 57)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 495)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 129)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 496)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 58)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 496)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 130)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 497)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 59)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 497)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 131)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 576)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 66)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 576)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 138)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 577)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 67)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 577)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 139)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 578)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 68)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 578)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 140)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 78)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 79)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 80)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 87)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 88)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 89)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 96)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 25)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 97)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 26)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 98)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 33)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 105)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 34)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 106)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 35)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 107)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 342)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 42)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 342)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 114)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 343)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 43)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 343)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 115)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 344)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 44)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 344)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 116)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 423)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 51)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 423)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 123)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 424)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 52)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 424)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 124)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 425)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 53)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 425)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 125)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 504)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 60)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 504)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 132)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 505)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 61)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 505)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 133)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 506)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 62)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 506)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 134)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 585)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 69)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 585)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 141)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 586)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 70)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 586)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 142)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 587)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 71)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 587)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 143)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[((((int)threadIdx.x) / 7) * 144)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 576)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1152)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1728)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 36)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 612)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1188)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1764)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 577)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1153)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1729)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 37)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 613)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1189)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1765)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 2)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 578)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1154)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1730)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 38)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 614)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1190)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1766)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 3)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 579)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1155)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1731)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 39)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 615)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1191)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1767)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 4)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 580)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1156)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1732)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 40)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 616)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1192)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1768)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 5)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 581)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1157)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1733)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 41)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 617)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1193)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1769)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 6)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 582)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1158)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1734)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 42)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 618)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1194)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1770)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 7)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 583)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1159)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1735)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 43)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 619)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1195)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1771)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 8)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 584)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1160)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1736)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 44)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 620)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1196)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1772)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 72)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 648)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1224)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1800)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 108)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 684)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1260)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1836)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 73)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 649)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1225)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1801)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 109)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 685)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1261)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1837)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 74)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 650)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1226)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1802)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 110)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 686)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1262)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1838)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 75)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 651)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1227)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1803)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 111)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 687)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1263)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1839)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 76)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 652)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1228)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1804)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 112)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 688)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1264)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1840)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 77)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 653)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1229)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1805)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 113)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 689)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1265)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1841)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 78)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 654)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1230)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1806)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 114)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 690)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1266)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1842)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 79)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 655)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1231)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1807)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 115)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 691)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1267)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1843)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 80)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 656)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1232)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1808)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 116)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 692)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1268)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1844)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 9)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 585)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1161)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1737)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 45)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 621)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1197)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1773)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 10)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 586)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1162)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1738)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 46)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 622)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1198)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1774)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 11)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 587)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1163)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1739)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 47)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 623)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1199)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1775)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 12)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 588)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1164)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1740)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 48)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 624)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1200)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1776)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 13)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 589)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1165)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1741)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 49)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 625)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1201)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1777)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 14)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 590)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1166)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1742)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 50)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 626)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1202)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1778)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 15)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 591)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1167)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1743)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 51)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 627)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1203)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1779)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 16)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 592)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1168)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1744)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 52)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 628)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1204)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1780)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 17)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 593)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1169)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1745)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 53)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 629)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1205)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1781)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 81)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 657)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1233)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1809)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 117)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 693)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1269)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1845)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 82)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 658)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1234)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1810)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 118)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 694)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1270)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1846)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 83)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 659)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1235)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1811)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 119)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 695)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1271)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1847)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 84)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 660)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1236)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1812)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 120)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 696)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1272)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1848)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 85)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 661)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1237)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1813)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 121)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 697)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1273)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1849)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 86)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 662)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1238)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1814)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 122)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 698)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1274)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1850)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 87)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 663)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1239)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1815)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 123)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 699)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1275)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1851)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 88)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 664)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1240)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1816)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 124)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 700)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1276)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1852)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 89)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 665)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1241)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1817)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 125)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 701)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1277)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1853)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 18)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 594)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1170)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1746)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 54)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 630)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1206)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1782)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 19)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 595)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1171)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1747)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 55)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 631)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1207)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1783)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 20)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 596)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1172)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1748)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 56)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 632)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1208)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1784)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 21)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 597)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1173)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1749)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 57)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 633)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1209)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1785)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 22)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 598)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1174)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1750)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 58)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 634)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1210)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1786)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 23)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 599)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1175)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1751)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 59)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 635)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1211)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1787)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 24)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 600)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1176)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1752)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 60)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 636)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1212)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1788)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 25)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 601)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1177)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1753)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 61)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 637)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1213)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1789)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 26)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 602)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1178)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1754)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 62)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 638)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1214)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1790)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 90)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 666)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1242)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1818)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 126)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 702)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1278)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1854)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 91)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 667)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1243)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1819)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 127)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 703)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1279)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1855)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 92)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 668)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1244)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1820)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 128)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 704)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1280)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1856)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 93)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 669)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1245)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1821)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 129)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 705)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1281)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1857)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 94)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 670)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1246)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1822)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 130)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 706)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1282)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1858)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 95)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 671)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1247)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1823)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 131)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 707)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1283)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1859)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 96)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 672)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1248)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1824)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 132)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 708)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1284)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1860)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 97)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 673)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1249)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1825)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 133)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 709)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1285)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1861)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 98)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 674)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1250)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1826)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 134)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 710)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1286)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1862)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 27)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 603)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1179)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1755)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 63)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 639)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1215)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1791)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 28)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 604)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1180)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1756)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 64)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 640)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1216)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1792)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 29)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 605)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1181)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1757)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 65)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 641)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1217)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1793)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 30)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 606)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1182)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1758)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 66)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 642)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1218)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1794)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 31)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 607)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1183)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1759)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 67)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 643)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1219)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1795)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 32)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 608)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1184)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1760)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 68)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 644)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1220)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1796)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 33)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 609)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1185)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1761)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 69)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 645)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1221)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1797)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 34)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 610)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1186)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1762)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 70)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 646)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1222)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1798)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 35)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 611)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1187)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1763)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 71)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 647)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1223)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1799)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 99)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 675)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1251)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1827)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 135)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 711)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1287)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1863)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 100)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 676)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1252)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1828)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 136)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 712)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1288)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1864)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 101)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 677)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1253)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1829)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 137)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 713)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1289)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1865)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 102)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 678)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1254)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1830)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 138)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 714)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1290)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1866)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 103)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 679)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1255)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1831)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 139)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 715)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1291)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1867)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 104)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 680)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1256)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1832)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 140)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 716)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1292)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1868)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 105)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 681)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1257)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1833)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 141)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 717)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1293)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1869)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 106)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 682)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1258)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1834)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 142)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 718)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1294)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1870)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 107)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 683)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1259)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1835)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 143)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 719)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1295)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1871)]));
}
- for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
- compute[((((((int)blockIdx.x) * 1568) + ((((int)threadIdx.x) / 49) * 98)) + (i1_inner * 49)) + (((int)threadIdx.x) % 49))] = max((conv2d_nchw[i1_inner] + bias[(((((int)blockIdx.x) * 32) + ((((int)threadIdx.x) / 49) * 2)) + i1_inner)]), 0.000000e+00f);
+ for (int i1_inner = 0; i1_inner < 4; ++i1_inner) {
+ compute[((((((((int)blockIdx.x) / 7) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 7))] = max((conv2d_nchw[i1_inner] + bias[((((((int)blockIdx.x) / 7) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner)]), 0.000000e+00f);
+ compute[(((((((((int)blockIdx.x) / 7) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 7)) + 784)] = max((conv2d_nchw[(i1_inner + 4)] + bias[(((((((int)blockIdx.x) / 7) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner) + 16)]), 0.000000e+00f);
+ compute[(((((((((int)blockIdx.x) / 7) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 7)) + 1568)] = max((conv2d_nchw[(i1_inner + 8)] + bias[(((((((int)blockIdx.x) / 7) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner) + 32)]), 0.000000e+00f);
+ compute[(((((((((int)blockIdx.x) / 7) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 7)) + 2352)] = max((conv2d_nchw[(i1_inner + 12)] + bias[(((((((int)blockIdx.x) / 7) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner) + 48)]), 0.000000e+00f);
}
}
@@ -815,7 +1963,7 @@ In the example below we resume the status and do more 5 trials.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 5 minutes 59.876 seconds)
+ **Total running time of the script:** ( 6 minutes 8.809 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 ad5418b941..b373bbd974 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt
@@ -647,7 +647,7 @@ so we can read the log file and load the best schedules.
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 7.9015 7.8961 7.9151 7.8933 0.0097
+ 7.9351 7.9334 7.9456 7.9262 0.0080
@@ -675,7 +675,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 7.619 seconds)
+ **Total running time of the script:** ( 1 minutes 7.165 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 d55c238f0f..2df2245bde 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)
- 748.8999 748.8494 749.0310 748.8193 0.0935
+ 760.6615 760.3228 762.8023 758.8594 1.6274
@@ -694,7 +694,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 40.136 seconds)
+ **Total running time of the script:** ( 1 minutes 40.129 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 4f7f0608a3..8fd26c514a 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
@@ -389,87 +389,404 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
@T.prim_func
def main(placeholder: T.Buffer((128, 256), "float32"), placeholder_1: T.Buffer((4916, 16, 1), "float32"), placeholder_2: T.Buffer((4916,), "int32"), placeholder_3: T.Buffer((33,), "int32"), placeholder_4: T.Buffer((128, 512), "float32"), compute: T.Buffer((128, 512), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
- for i0_outer in T.parallel(4):
- compute_1 = T.allocate([512], "float32", "global")
- for i1_outer in range(32):
- compute_2 = T.Buffer((512,), data=compute_1)
- for i_outer_inner in range(4):
- for i_inner_init in range(8):
- cse_var_1: T.int32 = i_outer_inner * 128 + i_inner_init * 16
- compute_2[cse_var_1] = T.float32(0)
- compute_2[cse_var_1 + 1] = T.float32(0)
- compute_2[cse_var_1 + 2] = T.float32(0)
- compute_2[cse_var_1 + 3] = T.float32(0)
- compute_2[cse_var_1 + 4] = T.float32(0)
- compute_2[cse_var_1 + 5] = T.float32(0)
- compute_2[cse_var_1 + 6] = T.float32(0)
- compute_2[cse_var_1 + 7] = T.float32(0)
- compute_2[cse_var_1 + 8] = T.float32(0)
- compute_2[cse_var_1 + 9] = T.float32(0)
- compute_2[cse_var_1 + 10] = T.float32(0)
- compute_2[cse_var_1 + 11] = T.float32(0)
- compute_2[cse_var_1 + 12] = T.float32(0)
- compute_2[cse_var_1 + 13] = T.float32(0)
- compute_2[cse_var_1 + 14] = T.float32(0)
- compute_2[cse_var_1 + 15] = T.float32(0)
- for elem_idx, i_inner in T.grid(placeholder_5[i1_outer + 1] - placeholder_5[i1_outer], 8):
- placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
- 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)
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_2: T.int32 = i_outer_inner * 128 + i_inner * 16
- compute_2[cse_var_2] = compute_2[cse_var_2] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_3: T.int32 = i_outer_inner * 128 + i_inner * 16 + 1
- compute_2[cse_var_3] = compute_2[cse_var_3] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_4: T.int32 = i_outer_inner * 128 + i_inner * 16 + 2
- compute_2[cse_var_4] = compute_2[cse_var_4] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_5: T.int32 = i_outer_inner * 128 + i_inner * 16 + 3
- compute_2[cse_var_5] = compute_2[cse_var_5] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_6: T.int32 = i_outer_inner * 128 + i_inner * 16 + 4
- compute_2[cse_var_6] = compute_2[cse_var_6] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_7: T.int32 = i_outer_inner * 128 + i_inner * 16 + 5
- compute_2[cse_var_7] = compute_2[cse_var_7] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_8: T.int32 = i_outer_inner * 128 + i_inner * 16 + 6
- compute_2[cse_var_8] = compute_2[cse_var_8] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_9: T.int32 = i_outer_inner * 128 + i_inner * 16 + 7
- compute_2[cse_var_9] = compute_2[cse_var_9] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_10: T.int32 = i_outer_inner * 128 + i_inner * 16 + 8
- compute_2[cse_var_10] = compute_2[cse_var_10] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_11: T.int32 = i_outer_inner * 128 + i_inner * 16 + 9
- compute_2[cse_var_11] = compute_2[cse_var_11] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_12: T.int32 = i_outer_inner * 128 + i_inner * 16 + 10
- compute_2[cse_var_12] = compute_2[cse_var_12] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_13: T.int32 = i_outer_inner * 128 + i_inner * 16 + 11
- compute_2[cse_var_13] = compute_2[cse_var_13] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_14: T.int32 = i_outer_inner * 128 + i_inner * 16 + 12
- compute_2[cse_var_14] = compute_2[cse_var_14] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_15: T.int32 = i_outer_inner * 128 + i_inner * 16 + 13
- compute_2[cse_var_15] = compute_2[cse_var_15] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_16: T.int32 = i_outer_inner * 128 + i_inner * 16 + 14
- compute_2[cse_var_16] = compute_2[cse_var_16] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_17: T.int32 = i_outer_inner * 128 + i_inner * 16 + 15
- compute_2[cse_var_17] = compute_2[cse_var_17] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- for i0_inner in range(32):
- cse_var_18: T.int32 = i0_outer * 16384 + i0_inner * 512 + i1_outer * 16
- compute_3 = T.Buffer((65536,), data=compute.data)
- placeholder_5 = T.Buffer((65536,), data=placeholder_4.data)
- compute_3[cse_var_18:cse_var_18 + 16] = T.max(compute_2[i0_inner * 16:i0_inner * 16 + 16] + placeholder_5[cse_var_18:cse_var_18 + 16], T.Broadcast(T.float32(0), 16))
+ for i0_outer_i1_outer_fused in T.parallel(512):
+ cse_var_1: T.int32 = i0_outer_i1_outer_fused % 32
+ compute_1 = T.allocate([128], "float32", "global")
+ compute_2 = T.Buffer((128,), data=compute_1)
+ compute_2[0] = T.float32(0)
+ compute_2[1] = T.float32(0)
+ compute_2[2] = T.float32(0)
+ compute_2[3] = T.float32(0)
+ compute_2[4] = T.float32(0)
+ compute_2[5] = T.float32(0)
+ compute_2[6] = T.float32(0)
+ compute_2[7] = T.float32(0)
+ compute_2[8] = T.float32(0)
+ compute_2[9] = T.float32(0)
+ compute_2[10] = T.float32(0)
+ compute_2[11] = T.float32(0)
+ compute_2[12] = T.float32(0)
+ compute_2[13] = T.float32(0)
+ compute_2[14] = T.float32(0)
+ compute_2[15] = T.float32(0)
+ compute_2[16] = T.float32(0)
+ compute_2[17] = T.float32(0)
+ compute_2[18] = T.float32(0)
+ compute_2[19] = T.float32(0)
+ compute_2[20] = T.float32(0)
+ compute_2[21] = T.float32(0)
+ compute_2[22] = T.float32(0)
+ compute_2[23] = T.float32(0)
+ compute_2[24] = T.float32(0)
+ compute_2[25] = T.float32(0)
+ compute_2[26] = T.float32(0)
+ compute_2[27] = T.float32(0)
+ compute_2[28] = T.float32(0)
+ compute_2[29] = T.float32(0)
+ compute_2[30] = T.float32(0)
+ compute_2[31] = T.float32(0)
+ compute_2[32] = T.float32(0)
+ compute_2[33] = T.float32(0)
+ compute_2[34] = T.float32(0)
+ compute_2[35] = T.float32(0)
+ compute_2[36] = T.float32(0)
+ compute_2[37] = T.float32(0)
+ compute_2[38] = T.float32(0)
+ compute_2[39] = T.float32(0)
+ compute_2[40] = T.float32(0)
+ compute_2[41] = T.float32(0)
+ compute_2[42] = T.float32(0)
+ compute_2[43] = T.float32(0)
+ compute_2[44] = T.float32(0)
+ compute_2[45] = T.float32(0)
+ compute_2[46] = T.float32(0)
+ compute_2[47] = T.float32(0)
+ compute_2[48] = T.float32(0)
+ compute_2[49] = T.float32(0)
+ compute_2[50] = T.float32(0)
+ compute_2[51] = T.float32(0)
+ compute_2[52] = T.float32(0)
+ compute_2[53] = T.float32(0)
+ compute_2[54] = T.float32(0)
+ compute_2[55] = T.float32(0)
+ compute_2[56] = T.float32(0)
+ compute_2[57] = T.float32(0)
+ compute_2[58] = T.float32(0)
+ compute_2[59] = T.float32(0)
+ compute_2[60] = T.float32(0)
+ compute_2[61] = T.float32(0)
+ compute_2[62] = T.float32(0)
+ compute_2[63] = T.float32(0)
+ compute_2[64] = T.float32(0)
+ compute_2[65] = T.float32(0)
+ compute_2[66] = T.float32(0)
+ compute_2[67] = T.float32(0)
+ compute_2[68] = T.float32(0)
+ compute_2[69] = T.float32(0)
+ compute_2[70] = T.float32(0)
+ compute_2[71] = T.float32(0)
+ compute_2[72] = T.float32(0)
+ compute_2[73] = T.float32(0)
+ compute_2[74] = T.float32(0)
+ compute_2[75] = T.float32(0)
+ compute_2[76] = T.float32(0)
+ compute_2[77] = T.float32(0)
+ compute_2[78] = T.float32(0)
+ compute_2[79] = T.float32(0)
+ compute_2[80] = T.float32(0)
+ compute_2[81] = T.float32(0)
+ compute_2[82] = T.float32(0)
+ compute_2[83] = T.float32(0)
+ compute_2[84] = T.float32(0)
+ compute_2[85] = T.float32(0)
+ compute_2[86] = T.float32(0)
+ compute_2[87] = T.float32(0)
+ compute_2[88] = T.float32(0)
+ compute_2[89] = T.float32(0)
+ compute_2[90] = T.float32(0)
+ compute_2[91] = T.float32(0)
+ compute_2[92] = T.float32(0)
+ compute_2[93] = T.float32(0)
+ compute_2[94] = T.float32(0)
+ compute_2[95] = T.float32(0)
+ compute_2[96] = T.float32(0)
+ compute_2[97] = T.float32(0)
+ compute_2[98] = T.float32(0)
+ compute_2[99] = T.float32(0)
+ compute_2[100] = T.float32(0)
+ compute_2[101] = T.float32(0)
+ compute_2[102] = T.float32(0)
+ compute_2[103] = T.float32(0)
+ compute_2[104] = T.float32(0)
+ compute_2[105] = T.float32(0)
+ compute_2[106] = T.float32(0)
+ compute_2[107] = T.float32(0)
+ compute_2[108] = T.float32(0)
+ compute_2[109] = T.float32(0)
+ compute_2[110] = T.float32(0)
+ compute_2[111] = T.float32(0)
+ compute_2[112] = T.float32(0)
+ compute_2[113] = T.float32(0)
+ compute_2[114] = T.float32(0)
+ compute_2[115] = T.float32(0)
+ compute_2[116] = T.float32(0)
+ compute_2[117] = T.float32(0)
+ compute_2[118] = T.float32(0)
+ compute_2[119] = T.float32(0)
+ compute_2[120] = T.float32(0)
+ compute_2[121] = T.float32(0)
+ compute_2[122] = T.float32(0)
+ compute_2[123] = T.float32(0)
+ compute_2[124] = T.float32(0)
+ compute_2[125] = T.float32(0)
+ compute_2[126] = T.float32(0)
+ compute_2[127] = T.float32(0)
+ for elem_idx in range(placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
+ 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)
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[0] = compute_2[0] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[1] = compute_2[1] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[2] = compute_2[2] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[3] = compute_2[3] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[4] = compute_2[4] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[5] = compute_2[5] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[6] = compute_2[6] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[7] = compute_2[7] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[8] = compute_2[8] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[9] = compute_2[9] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[10] = compute_2[10] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[11] = compute_2[11] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[12] = compute_2[12] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[13] = compute_2[13] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[14] = compute_2[14] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[15] = compute_2[15] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[16] = compute_2[16] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[17] = compute_2[17] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[18] = compute_2[18] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[19] = compute_2[19] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[20] = compute_2[20] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[21] = compute_2[21] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[22] = compute_2[22] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[23] = compute_2[23] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[24] = compute_2[24] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[25] = compute_2[25] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[26] = compute_2[26] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[27] = compute_2[27] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[28] = compute_2[28] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[29] = compute_2[29] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[30] = compute_2[30] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[31] = compute_2[31] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[32] = compute_2[32] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[33] = compute_2[33] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[34] = compute_2[34] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[35] = compute_2[35] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[36] = compute_2[36] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[37] = compute_2[37] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[38] = compute_2[38] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[39] = compute_2[39] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[40] = compute_2[40] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[41] = compute_2[41] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[42] = compute_2[42] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[43] = compute_2[43] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[44] = compute_2[44] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[45] = compute_2[45] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[46] = compute_2[46] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[47] = compute_2[47] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[48] = compute_2[48] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[49] = compute_2[49] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[50] = compute_2[50] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[51] = compute_2[51] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[52] = compute_2[52] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[53] = compute_2[53] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[54] = compute_2[54] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[55] = compute_2[55] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[56] = compute_2[56] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[57] = compute_2[57] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[58] = compute_2[58] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[59] = compute_2[59] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[60] = compute_2[60] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[61] = compute_2[61] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[62] = compute_2[62] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[63] = compute_2[63] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[64] = compute_2[64] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[65] = compute_2[65] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[66] = compute_2[66] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[67] = compute_2[67] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[68] = compute_2[68] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[69] = compute_2[69] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[70] = compute_2[70] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[71] = compute_2[71] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[72] = compute_2[72] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[73] = compute_2[73] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[74] = compute_2[74] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[75] = compute_2[75] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[76] = compute_2[76] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[77] = compute_2[77] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[78] = compute_2[78] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[79] = compute_2[79] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[80] = compute_2[80] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[81] = compute_2[81] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[82] = compute_2[82] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[83] = compute_2[83] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[84] = compute_2[84] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[85] = compute_2[85] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[86] = compute_2[86] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[87] = compute_2[87] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[88] = compute_2[88] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[89] = compute_2[89] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[90] = compute_2[90] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[91] = compute_2[91] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[92] = compute_2[92] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[93] = compute_2[93] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[94] = compute_2[94] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[95] = compute_2[95] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[96] = compute_2[96] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[97] = compute_2[97] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[98] = compute_2[98] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[99] = compute_2[99] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[100] = compute_2[100] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[101] = compute_2[101] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[102] = compute_2[102] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[103] = compute_2[103] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[104] = compute_2[104] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[105] = compute_2[105] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[106] = compute_2[106] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[107] = compute_2[107] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[108] = compute_2[108] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[109] = compute_2[109] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[110] = compute_2[110] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[111] = compute_2[111] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[112] = compute_2[112] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[113] = compute_2[113] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[114] = compute_2[114] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[115] = compute_2[115] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[116] = compute_2[116] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[117] = compute_2[117] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[118] = compute_2[118] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[119] = compute_2[119] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[120] = compute_2[120] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[121] = compute_2[121] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[122] = compute_2[122] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[123] = compute_2[123] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[124] = compute_2[124] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[125] = compute_2[125] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[126] = compute_2[126] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[127] = compute_2[127] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ for i0_inner in range(8):
+ cse_var_2: T.int32 = i0_outer_i1_outer_fused // 32 * 4096 + i0_inner * 512 + cse_var_1 * 16
+ compute_3 = T.Buffer((65536,), data=compute.data)
+ placeholder_5 = T.Buffer((65536,), data=placeholder_4.data)
+ compute_3[cse_var_2:cse_var_2 + 16] = T.max(compute_2[i0_inner * 16:i0_inner * 16 + 16] + placeholder_5[cse_var_2:cse_var_2 + 16], T.Broadcast(T.float32(0), 16))
@@ -519,7 +836,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 1.814 ms
+ Execution time of this operator: 2.847 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 af07105018..badc41acf7 100644
--- a/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
@@ -5,16 +5,16 @@
Computation times
=================
-**00:34.993** total execution time for **how_to_tune_with_autotvm** files:
+**00:38.201** 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``) | 00:34.960 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``) | 00:38.168 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``) | 00:00.019 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``) | 00:00.020 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_cuda.py` (``tune_relay_cuda.py``) | 00:00.005 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_arm.py` (``tune_relay_arm.py``) | 00:00.004 | 0.0 MB |
-+--------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_mobile_gpu.py` (``tune_relay_mobile_gpu.py``) | 00:00.004 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_arm.py` (``tune_relay_arm.py``) | 00:00.004 | 0.0 MB |
++--------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt b/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
index 2fc6b0e01f..03a09b355e 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, 2, 64, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3173634
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 1, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3233179
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, 4, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1007576
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 64]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,890541
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)
@@ -636,10 +636,9 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 2, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6952174
- No: 4 GFLOPS: 61.52/61.52 result: MeasureResult(costs=(0.003763109814814815,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.1692252159118652, timestamp=1675844779.5716734) [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8912925
- No: 5 GFLOPS: 13.13/61.52 result: MeasureResult(costs=(0.01763057566666667,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.6716628074645996, timestamp=1675844783.3626807) [('tile_f', [-1, 1, 2, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7463608
- No: 6 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 8, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1372349
+ No: 4 GFLOPS: 6.98/6.98 result: MeasureResult(costs=(0.03316484825,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.683012962341309, timestamp=1675876164.6765957) [('tile_f', [-1, 4, 2, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7559592
+ No: 5 GFLOPS: 0.00/6.98 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -761,8 +760,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 16, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6763057
- No: 7 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 256, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5485083
+ No: 6 GFLOPS: 0.00/6.98 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
@@ -884,8 +883,8 @@ 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, 4, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1130603
- No: 8 GFLOPS: 0.00/61.52 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, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2731221
+ No: 7 GFLOPS: 0.00/6.98 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
@@ -1007,9 +1006,9 @@ 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, 16, 8, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9282365
- No: 9 GFLOPS: 8.43/61.52 result: MeasureResult(costs=(0.027449855750000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.60951566696167, timestamp=1675844785.193633) [('tile_f', [-1, 2, 1, 64]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,431401
- No: 10 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 8, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 64, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10184759
+ No: 8 GFLOPS: 394.41/394.41 result: MeasureResult(costs=(0.0005869529289099526,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4507114887237549, timestamp=1675876167.2906568) [('tile_f', [-1, 1, 8, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2237647
+ No: 9 GFLOPS: 0.00/394.41 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1131,8 +1130,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 2, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7462074
- No: 11 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 4, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 64, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6118936
+ No: 10 GFLOPS: 0.00/394.41 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
@@ -1254,8 +1253,9 @@ 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, 16, 4, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 512]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1160179
- No: 12 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 256, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,6985273
+ No: 11 GFLOPS: 44.86/394.41 result: MeasureResult(costs=(0.00516085775,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.5173752307891846, timestamp=1675876168.9924073) [('tile_f', [-1, 2, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5079401
+ No: 12 GFLOPS: 0.00/394.41 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
@@ -1377,8 +1377,8 @@ 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, 1, 256]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5088157
- No: 13 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 256, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2353351
+ No: 13 GFLOPS: 0.00/394.41 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
@@ -1500,8 +1500,9 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 4, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5505651
- No: 14 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ 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, 7, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7696988
+ No: 14 GFLOPS: 8.30/394.41 result: MeasureResult(costs=(0.0279019525,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8869495391845703, timestamp=1675876174.5089574) [('tile_f', [-1, 4, 8, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6004981
+ No: 15 GFLOPS: 0.00/394.41 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
@@ -1623,8 +1624,9 @@ 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, 16, 8, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8404083
- No: 15 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 16, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 128, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2964846
+ No: 16 GFLOPS: 485.98/485.98 result: MeasureResult(costs=(0.0004763566886227545,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.412088871002197, timestamp=1675876175.5081959) [('tile_f', [-1, 1, 16, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8988134
+ No: 17 GFLOPS: 0.00/485.98 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
@@ -1746,8 +1748,8 @@ 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, 1, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7376001
- No: 16 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 2, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,358584
+ No: 18 GFLOPS: 0.00/485.98 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
@@ -1869,8 +1871,8 @@ 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, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 256, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6066597
- No: 17 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 64, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2802054
+ No: 19 GFLOPS: 0.00/485.98 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
@@ -1992,284 +1994,8 @@ 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, 4, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4585536
- No: 18 GFLOPS: 2.23/61.52 result: MeasureResult(costs=(0.10378480525,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.782615900039673, timestamp=1675844790.4885643) [('tile_f', [-1, 16, 1, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6649999
- No: 19 GFLOPS: 0.00/61.52 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 357, 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 171, 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: 0x00007fabf2bc9fa2
- 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:185
- 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:1617
- 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
- at ../include/tvm/runtime/packed_func.h:1217
- 1: Call
- at ../include/tvm/runtime/packed_func.h:1213
- 0: operator()
- at ../src/runtime/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, 1, 1, 512]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9877339
- No: 20 GFLOPS: 0.00/61.52 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:1730
- 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
- at ../include/tvm/runtime/packed_func.h:1670
- 19: run<>
- at ../include/tvm/runtime/packed_func.h:1630
- 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1645
- 13: operator()
- at ../src/driver/driver_api.cc:395
- 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
- at ../src/driver/driver_api.cc:381
- 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
- at ../src/driver/driver_api.cc:276
- 10: tvm::transform::Pass::operator()(tvm::IRModule) const
- at ../src/ir/transform.cc:258
- 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:274
- 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:451
- 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:274
- 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/tir/ir/transform.cc:100
- 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
- at ../include/tvm/runtime/packed_func.h:1749
- 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
- at ../include/tvm/runtime/packed_func.h:1693
- 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
- at ../include/tvm/runtime/packed_func.h:1617
- 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
- at ../include/tvm/runtime/packed_func.h:1217
- 1: Call
- at ../include/tvm/runtime/packed_func.h:1213
- 0: operator()
- at ../src/runtime/c_runtime_api.cc:534
- File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
- File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
- raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel
-
- Traceback (most recent call last):
- 24: TVMFuncCall
- at ../src/runtime/c_runtime_api.cc:477
- 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
- at ../include/tvm/runtime/packed_func.h:1217
- 22: Call
- at ../include/tvm/runtime/packed_func.h:1213
- 21: operator()
- at ../include/tvm/runtime/packed_func.h:1730
- 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
- at ../include/tvm/runtime/packed_func.h:1670
- 19: run<>
- at ../include/tvm/runtime/packed_func.h:1630
- 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1645
- 13: operator()
- at ../src/driver/driver_api.cc:395
- 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
- at ../src/driver/driver_api.cc:381
- 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
- at ../src/driver/driver_api.cc:276
- 10: tvm::transform::Pass::operator()(tvm::IRModule) const
- at ../src/ir/transform.cc:258
- 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:274
- 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:451
- 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:274
- 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/tir/ir/transform.cc:100
- 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
- at ../include/tvm/runtime/packed_func.h:1749
- 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
- at ../include/tvm/runtime/packed_func.h:1693
- 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
- at ../include/tvm/runtime/packed_func.h:1617
- 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
- at ../include/tvm/runtime/packed_func.h:1217
- 1: Call
- at ../include/tvm/runtime/packed_func.h:1213
- 0: operator()
- at ../src/runtime/c_runtime_api.cc:534
- File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
- File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
- raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 1, 2]), ('tile_y', [-1, 7, 1, 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', 512), ('unroll_explicit', 0)],None,2257481
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 32, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 8, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2016564
+ No: 20 GFLOPS: 253.61/485.98 result: MeasureResult(costs=(0.0009128193448275862,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.485121250152588, timestamp=1675876177.2099645) [('tile_f', [-1, 1, 8, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2419021
@@ -2324,9 +2050,9 @@ and measure running time.
Finish loading 20 records
Best config:
- [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8912925
+ [('tile_f', [-1, 1, 16, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8988134
Finish loading 20 records
- Time cost of this operator: 0.001784
+ Time cost of this operator: 0.000840
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 625fd5420c..25fb6854bb 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 311.2 98.689 (1, 2, 10, 10, 3) 2 1 [311.2]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.125 0.991 (1, 6, 10, 10) 1 1 [3.125]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 1.01 0.32 (1, 1, 10, 10, 3) 1 1 [1.01]
- Total_time - 315.335 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 308.5 98.68 (1, 2, 10, 10, 3) 2 1 [308.5]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.164 1.012 (1, 6, 10, 10) 1 1 [3.164]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.963 0.308 (1, 1, 10, 10, 3) 1 1 [0.963]
+ Total_time - 312.627 - - - - -
@@ -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 103.2 97.508 (1, 6, 10, 10, 1) 2 1 [103.2]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.794 1.695 (1, 6, 10, 10) 1 1 [1.794]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.843 0.797 (1, 3, 10, 10, 1) 1 1 [0.843]
- Total_time - 105.837 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 100.2 97.295 (1, 6, 10, 10, 1) 2 1 [100.2]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.816 1.763 (1, 6, 10, 10) 1 1 [1.816]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.97 0.942 (1, 1, 10, 10, 3) 1 1 [0.97]
+ Total_time - 102.986 - - - - -
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 4637ce754e..3b33619896 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, 43.7MB/s]
+
0%| | 0.00/3.42M [00:00<?, ?B/s]
61%|###### | 2.09M/3.42M [00:00<00:00, 13.0MB/s]
100%|##########| 3.42M/3.42M [00:00<00:00, 20.4MB/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 11.212 seconds)
+ **Total running time of the script:** ( 1 minutes 10.246 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 1c14eefabb..e5199fa5a5 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt
@@ -218,7 +218,7 @@ take about **2 minutes** to download the Stanford Cars, while COCO 2017 validati
.. code-block:: none
- '/tmp/tmpm50rqx90/images/random'
+ '/tmp/tmpy9cmecff/images/random'
@@ -309,7 +309,7 @@ objects to other stuff? We can display some examples from our datasets using ``m
.. image-sg:: /how_to/work_with_microtvm/images/sphx_glr_micro_train_001.png
- :alt: [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0]
+ :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
@@ -318,8 +318,8 @@ objects to other stuff? We can display some examples from our datasets using ``m
.. code-block:: none
- /tmp/tmpm50rqx90/images/target contains 8144 images
- /tmp/tmpm50rqx90/images/random contains 5000 images
+ /tmp/tmpy9cmecff/images/target contains 8144 images
+ /tmp/tmpy9cmecff/images/random contains 5000 images
@@ -494,13 +494,13 @@ the time on our validation set).
.. code-block:: none
Epoch 1/3
- 328/328 - 47s - loss: 0.2678 - accuracy: 0.9141 - val_loss: 0.1673 - val_accuracy: 0.9479 - 47s/epoch - 142ms/step
+ 328/328 - 47s - loss: 0.2250 - accuracy: 0.9227 - val_loss: 0.1374 - val_accuracy: 0.9528 - 47s/epoch - 144ms/step
Epoch 2/3
- 328/328 - 43s - loss: 0.1014 - accuracy: 0.9633 - val_loss: 0.1538 - val_accuracy: 0.9543 - 43s/epoch - 131ms/step
+ 328/328 - 44s - loss: 0.0961 - accuracy: 0.9655 - val_loss: 0.1376 - val_accuracy: 0.9569 - 44s/epoch - 133ms/step
Epoch 3/3
- 328/328 - 43s - loss: 0.0677 - accuracy: 0.9747 - val_loss: 0.1398 - val_accuracy: 0.9611 - 43s/epoch - 131ms/step
+ 328/328 - 44s - loss: 0.0702 - accuracy: 0.9747 - val_loss: 0.1168 - val_accuracy: 0.9603 - 44s/epoch - 133ms/step
- <keras.callbacks.History object at 0x7f1c024aad90>
+ <keras.callbacks.History object at 0x7efdc2500410>
@@ -861,7 +861,7 @@ 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:** ( 5 minutes 2.592 seconds)
+ **Total running time of the script:** ( 4 minutes 59.015 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 c83aabc1bb..cb8db4e9fd 100644
--- a/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
@@ -5,18 +5,18 @@
Computation times
=================
-**07:19.778** total execution time for **how_to_work_with_microtvm** files:
+**07:14.192** total execution time for **how_to_work_with_microtvm** files:
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 05:02.592 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:59.015 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``) | 01:11.212 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``) | 01:10.246 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 00:52.655 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 00:52.052 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:08.042 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:07.823 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:05.278 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:05.055 | 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 1f03efc933..9abffbb5f8 100644
--- a/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt
@@ -5,14 +5,14 @@
Computation times
=================
-**00:44.345** total execution time for **how_to_work_with_relay** files:
+**00:44.387** 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``) | 00:32.208 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:32.249 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:10.584 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:10.435 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.547 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.697 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``) | 00:00.006 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt b/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
index 8090a08bda..d1c37e0d49 100644
--- a/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
@@ -264,7 +264,7 @@ The following example customizes CUDA lowering rule for :code:`exp`.
.. code-block:: none
- <function my_cuda_math_rule at 0x7f1aaf1514d0>
+ <function my_cuda_math_rule at 0x7efc6c911560>
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 fe9dc6efc3..2b0c5b5553 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
=================
-**00:06.622** total execution time for **how_to_work_with_schedules** files:
+**00:08.952** total execution time for **how_to_work_with_schedules** files:
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:04.061 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:06.420 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``) | 00:01.195 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``) | 00:01.177 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.576 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.577 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.559 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.552 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.118 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.115 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.053 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``) | 00:00.033 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``) | 00:00.032 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``) | 00:00.027 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
index 8c1cd6fdc3..b9fe35ea1f 100644
--- a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
@@ -340,7 +340,7 @@ The importing needs to happen before the tensorized GEMV being executed.
def main(A: T.Buffer((1024, 64), "float32"), B: T.Buffer((512, 64), "float32"), C: T.Buffer((1024, 512), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
i = T.var("int32")
- T.attr(T.iter_var(i, None, "DataPar", ""), "pragma_import_llvm", "; ModuleID = '/tmp/tmpzoz5ee0i/input0.cc'\nsource_filename = \"/tmp/tmpzoz5ee0i/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = alloca float*, align 8\n %8 = alloca float*, align 8\n %9 = alloca float*, [...]
+ T.attr(T.iter_var(i, None, "DataPar", ""), "pragma_import_llvm", "; ModuleID = '/tmp/tmpxfbq4px4/input0.cc'\nsource_filename = \"/tmp/tmpxfbq4px4/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = alloca float*, align 8\n %8 = alloca float*, align 8\n %9 = alloca float*, [...]
for i, j_outer in T.grid(1024, 32):
T.call_extern("int32", "gemv_update", T.tvm_access_ptr(T.type_annotation("float32"), C.data, i * 512 + j_outer * 16, 16, 2), T.tvm_access_ptr(T.type_annotation("float32"), A.data, i * 64, 64, 1), T.tvm_access_ptr(T.type_annotation("float32"), B.data, j_outer * 1024, 1024, 1), 16, 64, 64)
diff --git a/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt
index cb5def7e0e..e179b3b217 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
=================
-**00:30.329** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:30.320** total execution time for **topic_vta_tutorials_autotvm** files:
+---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:30.322 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:30.314 | 0.0 MB |
+---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``) | 00:00.007 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``) | 00:00.006 | 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 efab8ddb52..11a3594cd8 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 32.33s!
+ resnet18_v1 inference graph built in 32.25s!
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 ba4d56bc9c..b6ad67a2a7 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 21.98s!
+ yolov3-tiny inference graph built in 22.06s!
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 e431608b11..6969bf6388 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
=================
-**01:38.586** total execution time for **topic_vta_tutorials_frontend** files:
+**01:38.171** total execution time for **topic_vta_tutorials_frontend** files:
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:49.439 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:49.156 | 0.0 MB |
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:49.147 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:49.014 | 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 9f8a8ce489..8355a988a5 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
=================
-**00:03.120** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.207** total execution time for **topic_vta_tutorials_optimize** files:
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.667 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.747 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.453 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.460 | 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 34d1b1cbb7..623ec5fd74 100644
--- a/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
@@ -8,7 +8,7 @@ Computation times
**00:00.769** total execution time for **topic_vta_tutorials** files:
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.400 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.398 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.369 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.371 | 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 9fa4e4dee0..bd72f7868f 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: 95.769 ms
+ Execution time of this operator: 98.051 ms
@@ -436,7 +436,7 @@ operations.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 32.780 seconds)
+ **Total running time of the script:** ( 1 minutes 21.768 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 c2fd34cd8f..68d56e84d4 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: 3.70/3.70 result: MeasureResult(costs=(0.072537796,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4144327640533447, timestamp=1675843202.9258363) [('tile_y', [-1, 128]), ('tile_x', [-1, 16])],None,47
- No: 2 GFLOPS: 2.43/3.70 result: MeasureResult(costs=(0.11053074739999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.003974199295044, timestamp=1675843205.7111332) [('tile_y', [-1, 2]), ('tile_x', [-1, 4])],None,21
- No: 3 GFLOPS: 13.88/13.88 result: MeasureResult(costs=(0.019345932,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5521788597106934, timestamp=1675843206.2700112) [('tile_y', [-1, 128]), ('tile_x', [-1, 64])],None,67
- No: 4 GFLOPS: 2.13/13.88 result: MeasureResult(costs=(0.1258743678,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.2409310340881348, timestamp=1675843209.299456) [('tile_y', [-1, 1]), ('tile_x', [-1, 8])],None,30
- No: 5 GFLOPS: 12.00/13.88 result: MeasureResult(costs=(0.0223702194,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5988986492156982, timestamp=1675843210.326542) [('tile_y', [-1, 256]), ('tile_x', [-1, 256])],None,88
- No: 6 GFLOPS: 2.84/13.88 result: MeasureResult(costs=(0.0944511938,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7525465488433838, timestamp=1675843212.8573265) [('tile_y', [-1, 16]), ('tile_x', [-1, 4])],None,24
- No: 7 GFLOPS: 14.80/14.80 result: MeasureResult(costs=(0.0181364724,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.829948902130127, timestamp=1675843213.3957038) [('tile_y', [-1, 64]), ('tile_x', [-1, 64])],None,66
- No: 8 GFLOPS: 2.91/14.80 result: MeasureResult(costs=(0.0922643292,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7266316413879395, timestamp=1675843215.1203144) [('tile_y', [-1, 2]), ('tile_x', [-1, 16])],None,41
- No: 9 GFLOPS: 12.72/14.80 result: MeasureResult(costs=(0.021100753,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5546364784240723, timestamp=1675843215.7911093) [('tile_y', [-1, 32]), ('tile_x', [-1, 128])],None,75
- No: 10 GFLOPS: 2.82/14.80 result: MeasureResult(costs=(0.0952459448,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7294600009918213, timestamp=1675843217.5656853) [('tile_y', [-1, 1]), ('tile_x', [-1, 16])],None,40
+ No: 1 GFLOPS: 13.62/13.62 result: MeasureResult(costs=(0.0197147112,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5657351016998291, timestamp=1675874608.8638072) [('tile_y', [-1, 256]), ('tile_x', [-1, 64])],None,68
+ No: 2 GFLOPS: 12.96/13.62 result: MeasureResult(costs=(0.0207108036,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6452291011810303, timestamp=1675874609.4477046) [('tile_y', [-1, 128]), ('tile_x', [-1, 128])],None,77
+ No: 3 GFLOPS: 11.28/13.62 result: MeasureResult(costs=(0.023796493000000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.60581374168396, timestamp=1675874610.826277) [('tile_y', [-1, 512]), ('tile_x', [-1, 512])],None,99
+ No: 4 GFLOPS: 11.90/13.62 result: MeasureResult(costs=(0.022562949000000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5879402160644531, timestamp=1675874611.4419045) [('tile_y', [-1, 2]), ('tile_x', [-1, 512])],None,91
+ No: 5 GFLOPS: 1.53/13.62 result: MeasureResult(costs=(0.1756962172,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.044736862182617, timestamp=1675874614.6117253) [('tile_y', [-1, 64]), ('tile_x', [-1, 4])],None,26
+ No: 6 GFLOPS: 2.27/13.62 result: MeasureResult(costs=(0.11839204380000001,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.135852813720703, timestamp=1675874617.5071929) [('tile_y', [-1, 4]), ('tile_x', [-1, 2])],None,12
+ No: 7 GFLOPS: 3.69/13.62 result: MeasureResult(costs=(0.0726688388,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4000210762023926, timestamp=1675874619.670283) [('tile_y', [-1, 16]), ('tile_x', [-1, 8])],None,34
+ No: 8 GFLOPS: 2.99/13.62 result: MeasureResult(costs=(0.08968996879999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.681145429611206, timestamp=1675874621.3546863) [('tile_y', [-1, 2]), ('tile_x', [-1, 16])],None,41
+ No: 9 GFLOPS: 2.11/13.62 result: MeasureResult(costs=(0.1270647646,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.2380712032318115, timestamp=1675874623.706845) [('tile_y', [-1, 8]), ('tile_x', [-1, 2])],None,13
+ No: 10 GFLOPS: 10.52/13.62 result: MeasureResult(costs=(0.0255153454,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6043596267700195, timestamp=1675874624.3679667) [('tile_y', [-1, 1]), ('tile_x', [-1, 512])],None,90
diff --git a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
index fb074ff87f..d590f13378 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': 514.7913624499994, 'median': 514.5572220499957, 'std': 1.832343611061247}
+ {'mean': 524.7727405299996, 'median': 524.005764550003, 'std': 2.8968953730024656}
@@ -545,31 +545,30 @@ the tuning data to.
.. code-block:: none
-
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 12.02/ 13.01 GFLOPS | Progress: (4/20) | 8.51 s
[Task 1/25] Current/Best: 6.79/ 23.09 GFLOPS | Progress: (8/20) | 11.16 s
[Task 1/25] Current/Best: 15.70/ 23.13 GFLOPS | Progress: (12/20) | 12.98 s
[Task 1/25] Current/Best: 7.08/ 23.13 GFLOPS | Progress: (16/20) | 16.89 s
[Task 1/25] Current/Best: 19.50/ 23.13 GFLOPS | Progress: (20/20) | 19.46 s Done.
-
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 18.74/ 18.74 GFLOPS | Progress: (4/20) | 3.24 s
[Task 2/25] Current/Best: 3.51/ 18.74 GFLOPS | Progress: (8/20) | 6.20 s
[Task 2/25] Current/Best: 6.36/ 18.74 GFLOPS | Progress: (12/20) | 7.66 s
[Task 2/25] Current/Best: 11.76/ 18.74 GFLOPS | Progress: (16/20) | 9.42 s
[Task 2/25] Current/Best: 19.74/ 19.74 GFLOPS | Progress: (20/20) | 10.90 s Done.
-
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 23.24/ 23.24 GFLOPS | Progress: (4/20) | 3.74 s
[Task 3/25] Current/Best: 6.86/ 23.24 GFLOPS | Progress: (8/20) | 6.51 s
[Task 3/25] Current/Best: 12.08/ 23.24 GFLOPS | Progress: (12/20) | 9.23 s
[Task 3/25] Current/Best: 5.68/ 23.24 GFLOPS | Progress: (16/20) | 11.67 s
[Task 3/25] Current/Best: 18.73/ 23.24 GFLOPS | Progress: (20/20) | 13.45 s Done.
-
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 12.76/ 16.65 GFLOPS | Progress: (4/20) | 4.79 s
[Task 4/25] Current/Best: 13.89/ 20.39 GFLOPS | Progress: (8/20) | 7.17 s
[Task 4/25] Current/Best: 4.72/ 20.39 GFLOPS | Progress: (12/20) | 14.31 s
[Task 4/25] Current/Best: 17.99/ 20.39 GFLOPS | Progress: (16/20) | 16.01 s
[Task 4/25] Current/Best: 9.94/ 20.39 GFLOPS | Progress: (20/20) | 18.35 s Done.
-
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 10.15/ 10.15 GFLOPS | Progress: (4/20) | 4.01 s
[Task 5/25] Current/Best: 10.62/ 14.94 GFLOPS | Progress: (8/20) | 7.51 s
[Task 5/25] Current/Best: 3.33/ 19.89 GFLOPS | Progress: (12/20) | 9.61 s
[Task 5/25] Current/Best: 12.23/ 19.89 GFLOPS | Progress: (16/20) | 11.76 s
[Task 5/25] Current/Best: 7.01/ 19.89 GFLOPS | Progress: (20/20) | 13.74 s Done.
-
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 18.94/ 18.94 GFLOPS | Progress: (4/20) | 4.18 s
[Task 6/25] Current/Best: 16.00/ 18.94 GFLOPS | Progress: (8/20) | 6.28 s
[Task 6/25] Current/Best: 9.18/ 19.56 GFLOPS | Progress: (12/20) | 8.82 s
[Task 6/25] Current/Best: 13.15/ 19.56 GFLOPS | Progress: (16/20) | 11.31 s
[Task 6/25] Current/Best: 18.09/ 21.75 GFLOPS | Progress: (20/20) | 13.33 s Done.
-
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 7.65/ 12.95 GFLOPS | Progress: (4/20) | 4.12 s
[Task 7/25] Current/Best: 14.97/ 18.41 GFLOPS | Progress: (8/20) | 6.54 s
[Task 7/25] Current/Best: 12.74/ 18.92 GFLOPS | Progress: (12/20) | 8.69 s
[Task 7/25] Current/Best: 20.64/ 20.64 GFLOPS | Progress: (16/20) | 11.39 s
[Task 7/25] Current/Best: 3.09/ 20.64 GFLOPS | Progress: (20/20) | 14.93 s Done.
-
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 19.34/ 19.34 GFLOPS | Progress: (4/20) | 10.19 s
[Task 8/25] Current/Best: 11.62/ 19.34 GFLOPS | Progress: (8/20) | 14.13 s
[Task 8/25] Current/Best: 13.94/ 19.34 GFLOPS | Progress: (12/20) | 22.49 s
[Task 8/25] Current/Best: 8.66/ 19.34 GFLOPS | Progress: (16/20) | 26.42 s
[Task 8/25] Current/Best: 2.77/ 19.34 GFLOPS | Progress: (20/20) | 36.92 s Done.
-
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 5.22/ 19.38 GFLOPS | Progress: (4/20) | 7.41 s
[Task 9/25] Current/Best: 14.39/ 19.38 GFLOPS | Progress: (8/20) | 18.46 s
[Task 9/25] Current/Best: 13.72/ 19.38 GFLOPS | Progress: (12/20) | 20.80 s
[Task 9/25] Current/Best: 16.96/ 21.83 GFLOPS | Progress: (16/20) | 31.77 s
[Task 9/25] Current/Best: 18.12/ 21.83 GFLOPS | Progress: (20/20) | 35.02 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 15.01/ 15.01 GFLOPS | Progress: (4/20) | 3.70 s
[Task 10/25] Current/Best: 11.24/ 15.01 GFLOPS | Progress: (8/20) | 5.65 s
[Task 10/25] Current/Best: 9.22/ 15.01 GFLOPS | Progress: (12/20) | 9.18 s
[Task 10/25] Current/Best: 2.95/ 17.65 GFLOPS | Progress: (16/20) | 12.02 s
[Task 10/25] Current/Best: 13.39/ 17.65 GFLOPS | Progress: (20/20
) | 13.84 s Done.
-
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 6.29/ 17.73 GFLOPS | Progress: (4/20) | 4.19 s
[Task 11/25] Current/Best: 1.58/ 19.44 GFLOPS | Progress: (8/20) | 8.11 s
[Task 11/25] Current/Best: 12.94/ 19.44 GFLOPS | Progress: (12/20) | 10.90 s
[Task 11/25] Current/Best: 22.87/ 22.87 GFLOPS | Progress: (16/20) | 13.36 s
[Task 11/25] Current/Best: 18.09/ 22.87 GFLOPS | Progress: (20/20) | 17.15 s Done.
-
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 15.75/ 15.75 GFLOPS | Progress: (4/20) | 4.96 s
[Task 12/25] Current/Best: 9.13/ 15.75 GFLOPS | Progress: (8/20) | 7.22 s
[Task 12/25] Current/Best: 11.52/ 15.80 GFLOPS | Progress: (12/20) | 9.27 s
[Task 12/25] Current/Best: 5.19/ 18.08 GFLOPS | Progress: (16/20) | 15.70 s
[Task 12/25] Current/Best: 10.47/ 18.55 GFLOPS | Progress: (20/20) | 20.29 s Done.
-
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 10.24/ 19.98 GFLOPS | Progress: (4/20) | 4.57 s
[Task 13/25] Current/Best: 7.01/ 19.98 GFLOPS | Progress: (8/20) | 8.23 s
[Task 13/25] Current/Best: 1.57/ 19.98 GFLOPS | Progress: (12/20) | 12.27 s
[Task 13/25] Current/Best: 13.89/ 19.98 GFLOPS | Progress: (16/20) | 14.59 s
[Task 13/25] Current/Best: 12.14/ 19.98 GFLOPS | Progress: (20/20) | 17.50 s Done.
-
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 5.80/ 12.48 GFLOPS | Progress: (4/20) | 3.78 s
[Task 14/25] Current/Best: 15.97/ 15.97 GFLOPS | Progress: (8/20) | 9.92 s Done.
-
[Task 14/25] Current/Best: 14.26/ 18.79 GFLOPS | Progress: (12/20) | 12.67 s
[Task 14/25] Current/Best: 8.61/ 18.79 GFLOPS | Progress: (16/20) | 16.05 s
[Task 14/25] Current/Best: 5.07/ 18.79 GFLOPS | Progress: (20/20) | 23.63 s Done.
-
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 13.05/ 13.29 GFLOPS | Progress: (4/20) | 3.42 s
[Task 15/25] Current/Best: 8.05/ 13.29 GFLOPS | Progress: (8/20) | 7.23 s
[Task 15/25] Current/Best: 6.32/ 20.33 GFLOPS | Progress: (12/20) | 8.88 s
[Task 15/25] Current/Best: 12.51/ 20.33 GFLOPS | Progress: (16/20) | 10.48 s
[Task 15/25] Current/Best: 20.22/ 22.37 GFLOPS | Progress: (20/20) | 14.27 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 18.45/ 18.64 GFLOPS | Progress: (4/20) | 4.26 s
[Task 16/25] Current/Best: 3.03/ 18.64 GFLOPS | Progress: (8/20) | 6.49 s
[Task 16/25] Current/Best: 5.38/ 18.64 GFLOPS | Progress: (12/20) | 8.75 s
[Task 16/25] Current/Best: 18.98/ 18.98 GFLOPS | Progress: (16/20) | 10.56 s
[Task 16/25] Current/Best: 6.52/ 20.31 GFLOPS | Progress: (20/20)
| 12.12 s Done.
-
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 11.88/ 13.74 GFLOPS | Progress: (4/20) | 5.27 s
[Task 17/25] Current/Best: 13.50/ 18.24 GFLOPS | Progress: (8/20) | 7.72 s
[Task 17/25] Current/Best: 15.18/ 18.53 GFLOPS | Progress: (12/20) | 9.99 s
[Task 17/25] Current/Best: 16.45/ 23.68 GFLOPS | Progress: (16/20) | 12.29 s
[Task 17/25] Current/Best: 19.03/ 23.68 GFLOPS | Progress: (20/20) | 14.94 s Done.
-
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 18.93/ 18.93 GFLOPS | Progress: (4/20) | 4.74 s
[Task 18/25] Current/Best: 16.36/ 18.93 GFLOPS | Progress: (8/20) | 11.33 s
[Task 18/25] Current/Best: 7.58/ 18.93 GFLOPS | Progress: (12/20) | 15.11 s
[Task 18/25] Current/Best: 5.01/ 18.93 GFLOPS | Progress: (16/20) | 17.58 s
[Task 18/25] Current/Best: 16.93/ 18.93 GFLOPS | Progress: (20/20) | 19.72 s Done.
-
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 7.86/ 21.40 GFLOPS | Progress: (4/20) | 5.67 s
[Task 19/25] Current/Best: 13.03/ 21.40 GFLOPS | Progress: (8/20) | 8.26 s
[Task 19/25] Current/Best: 20.25/ 21.40 GFLOPS | Progress: (12/20) | 12.00 s
[Task 19/25] Current/Best: 5.34/ 21.40 GFLOPS | Progress: (16/20) | 16.23 s
[Task 19/25] Current/Best: 16.44/ 21.40 GFLOPS | Progress: (20/20) | 19.63 s Done.
-
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 17.07/ 18.42 GFLOPS | Progress: (4/20) | 5.14 s
[Task 20/25] Current/Best: 6.24/ 18.42 GFLOPS | Progress: (8/20) | 7.67 s
[Task 20/25] Current/Best: 9.04/ 18.42 GFLOPS | Progress: (12/20) | 9.81 s
[Task 20/25] Current/Best: 2.53/ 18.42 GFLOPS | Progress: (16/20) | 16.03 s
[Task 20/25] Current/Best: 19.78/ 19.78 GFLOPS | Progress: (20/20) | 18.16 s Done.
-
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 11.96/ 11.96 GFLOPS | Progress: (4/20) | 3.34 s
[Task 21/25] Current/Best: 14.00/ 14.39 GFLOPS | Progress: (8/20) | 8.10 s
[Task 21/25] Current/Best: 10.66/ 18.82 GFLOPS | Progress: (12/20) | 10.34 s
[Task 21/25] Current/Best: 16.51/ 18.82 GFLOPS | Progress: (16/20) | 14.06 s
[Task 21/25] Current/Best: 9.78/ 18.82 GFLOPS | Progress: (20/20) | 15.77 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 15.62/ 15.62 GFLOPS | Progress: (4/20) | 3.76 s
[Task 22/25] Current/Best: 7.00/ 18.72 GFLOPS | Progress: (8/20) | 6.43 s Done.
+
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 21.97/ 23.00 GFLOPS | Progress: (4/20) | 7.83 s
[Task 1/25] Current/Best: 14.77/ 23.00 GFLOPS | Progress: (8/20) | 10.87 s
[Task 1/25] Current/Best: 10.42/ 23.00 GFLOPS | Progress: (12/20) | 15.21 s
[Task 1/25] Current/Best: 14.56/ 23.00 GFLOPS | Progress: (16/20) | 18.58 s
[Task 1/25] Current/Best: 10.59/ 24.09 GFLOPS | Progress: (20/20) | 20.88 s Done.
+
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 5.04/ 18.67 GFLOPS | Progress: (4/20) | 3.36 s
[Task 2/25] Current/Best: 15.73/ 18.67 GFLOPS | Progress: (8/20) | 5.18 s
[Task 2/25] Current/Best: 16.06/ 18.67 GFLOPS | Progress: (12/20) | 6.94 s
[Task 2/25] Current/Best: 16.81/ 18.67 GFLOPS | Progress: (16/20) | 8.89 s
[Task 2/25] Current/Best: 6.00/ 18.67 GFLOPS | Progress: (20/20) | 10.77 s Done.
+
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 12.95/ 23.85 GFLOPS | Progress: (4/20) | 3.85 s
[Task 3/25] Current/Best: 16.53/ 23.85 GFLOPS | Progress: (8/20) | 7.35 s
[Task 3/25] Current/Best: 15.40/ 23.85 GFLOPS | Progress: (12/20) | 9.24 s
[Task 3/25] Current/Best: 12.42/ 23.85 GFLOPS | Progress: (16/20) | 11.60 s
[Task 3/25] Current/Best: 16.38/ 23.85 GFLOPS | Progress: (20/20) | 14.39 s Done.
+
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 4.72/ 20.17 GFLOPS | Progress: (4/20) | 9.01 s
[Task 4/25] Current/Best: 9.15/ 20.17 GFLOPS | Progress: (8/20) | 17.91 s
[Task 4/25] Current/Best: 9.99/ 21.37 GFLOPS | Progress: (12/20) | 19.98 s
[Task 4/25] Current/Best: 20.04/ 21.37 GFLOPS | Progress: (16/20) | 22.26 s
[Task 4/25] Current/Best: 18.49/ 21.37 GFLOPS | Progress: (20/20) | 24.59 s Done.
+
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 12.29/ 20.17 GFLOPS | Progress: (4/20) | 4.19 s
[Task 5/25] Current/Best: 12.14/ 20.17 GFLOPS | Progress: (8/20) | 6.36 s
[Task 5/25] Current/Best: 20.94/ 20.94 GFLOPS | Progress: (12/20) | 8.32 s
[Task 5/25] Current/Best: 16.16/ 20.94 GFLOPS | Progress: (16/20) | 10.42 s
[Task 5/25] Current/Best: 18.00/ 20.94 GFLOPS | Progress: (20/20) | 12.01 s Done.
+
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 14.41/ 22.32 GFLOPS | Progress: (4/20) | 3.94 s
[Task 6/25] Current/Best: 3.04/ 22.32 GFLOPS | Progress: (8/20) | 6.67 s
[Task 6/25] Current/Best: 5.94/ 22.32 GFLOPS | Progress: (12/20) | 9.69 s
[Task 6/25] Current/Best: 10.22/ 22.32 GFLOPS | Progress: (16/20) | 16.49 s
[Task 6/25] Current/Best: 16.99/ 22.32 GFLOPS | Progress: (20/20) | 19.06 s Done.
+
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 15.54/ 15.54 GFLOPS | Progress: (4/20) | 4.29 s
[Task 7/25] Current/Best: 18.80/ 21.02 GFLOPS | Progress: (8/20) | 6.23 s
[Task 7/25] Current/Best: 13.90/ 21.02 GFLOPS | Progress: (12/20) | 8.59 s
[Task 7/25] Current/Best: 14.53/ 21.02 GFLOPS | Progress: (16/20) | 10.82 s
[Task 7/25] Current/Best: 11.80/ 21.02 GFLOPS | Progress: (20/20) | 13.37 s Done.
+
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 11.28/ 21.11 GFLOPS | Progress: (4/20) | 5.85 s
[Task 8/25] Current/Best: 19.14/ 21.11 GFLOPS | Progress: (8/20) | 8.20 s
[Task 8/25] Current/Best: 10.20/ 21.11 GFLOPS | Progress: (12/20) | 20.19 s
[Task 8/25] Current/Best: 9.86/ 21.11 GFLOPS | Progress: (16/20) | 25.90 s
[Task 8/25] Current/Best: 15.56/ 21.11 GFLOPS | Progress: (20/20) | 30.47 s
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 9.84/ 18.43 GFLOPS | Progress: (4/20) | 3.95 s
[Task 9/25] Current/Best: 19.48/ 20.55 GFLOPS | Progress: (8/20) | 11.45 s
[Task 9/25] Current/Best: 6.11/ 20.55 GFLOPS | Progress: (12/20) | 13.19 s
[Task 9/25] Current/Best: 19.90/ 20.55 GFLOPS | Progress: (16/20) | 15.15 s
[Task 9/25] Current/Best: 13.52/ 20.55 GFLOPS | Progress: (20/2
0) | 17.12 s Done.
+
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 8.02/ 13.84 GFLOPS | Progress: (4/20) | 5.43 s
[Task 10/25] Current/Best: 11.78/ 18.35 GFLOPS | Progress: (8/20) | 7.39 s
[Task 10/25] Current/Best: 8.85/ 18.35 GFLOPS | Progress: (12/20) | 9.45 s
[Task 10/25] Current/Best: 17.81/ 19.61 GFLOPS | Progress: (16/20) | 12.32 s
[Task 10/25] Current/Best: 10.41/ 19.61 GFLOPS | Progress: (20/20) | 14.34 s Done.
+
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 10.65/ 16.77 GFLOPS | Progress: (4/20) | 4.38 s
[Task 11/25] Current/Best: 9.30/ 16.77 GFLOPS | Progress: (8/20) | 6.95 s
[Task 11/25] Current/Best: 10.54/ 19.39 GFLOPS | Progress: (12/20) | 9.76 s
[Task 11/25] Current/Best: 10.64/ 19.39 GFLOPS | Progress: (16/20) | 12.64 s
[Task 11/25] Current/Best: 6.82/ 21.66 GFLOPS | Progress: (20/20) | 15.43 s Done.
+
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 11.17/ 21.80 GFLOPS | Progress: (4/20) | 3.83 s
[Task 12/25] Current/Best: 13.86/ 21.80 GFLOPS | Progress: (8/20) | 5.82 s
[Task 12/25] Current/Best: 12.93/ 21.80 GFLOPS | Progress: (12/20) | 9.80 s
[Task 12/25] Current/Best: 15.88/ 21.80 GFLOPS | Progress: (16/20) | 11.81 s
[Task 12/25] Current/Best: 10.80/ 21.80 GFLOPS | Progress: (20/20) | 17.58 s Done.
+
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 6.21/ 12.00 GFLOPS | Progress: (4/20) | 6.11 s
[Task 13/25] Current/Best: 13.11/ 17.68 GFLOPS | Progress: (8/20) | 8.70 s
[Task 13/25] Current/Best: 17.46/ 17.68 GFLOPS | Progress: (12/20) | 13.79 s
[Task 13/25] Current/Best: 18.29/ 18.29 GFLOPS | Progress: (16/20) | 17.30 s
[Task 13/25] Current/Best: 17.76/ 20.00 GFLOPS | Progress: (20/20) | 20.43 s Done.
+
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 14.55/ 19.01 GFLOPS | Progress: (4/20) | 3.57 s
[Task 14/25] Current/Best: 7.39/ 19.33 GFLOPS | Progress: (8/20) | 5.60 s
[Task 14/25] Current/Best: 14.97/ 19.33 GFLOPS | Progress: (12/20) | 9.66 s
[Task 14/25] Current/Best: 1.62/ 19.33 GFLOPS | Progress: (16/20) | 14.87 s
[Task 14/25] Current/Best: 13.41/ 19.33 GFLOPS | Progress: (20/20) | 16.99 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 21.01/ 21.01 GFLOPS | Progress: (4/20) | 6.83 s
[Task 15/25] Current/Best: 14.06/ 21.01 GFLOPS | Progress: (8/20) | 9.45 s
[Task 15/25] Current/Best: 15.72/ 21.01 GFLOPS | Progress: (12/20) | 11.76 s
[Task 15/25] Current/Best: 7.99/ 21.01 GFLOPS | Progress: (16/20) | 18.80 s
[Task 15/25] Current/Best: 12.12/ 21.01 GFLOPS | Progress: (20/20)
| 20.75 s Done.
+
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 15.63/ 18.32 GFLOPS | Progress: (4/20) | 4.10 s
[Task 16/25] Current/Best: 12.95/ 20.11 GFLOPS | Progress: (8/20) | 5.89 s
[Task 16/25] Current/Best: 4.20/ 20.11 GFLOPS | Progress: (12/20) | 7.65 s
[Task 16/25] Current/Best: 12.06/ 21.62 GFLOPS | Progress: (16/20) | 9.84 s
[Task 16/25] Current/Best: 14.31/ 21.62 GFLOPS | Progress: (20/20) | 11.85 s Done.
+
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 10.30/ 21.74 GFLOPS | Progress: (4/20) | 3.96 s
[Task 17/25] Current/Best: 13.05/ 22.54 GFLOPS | Progress: (8/20) | 6.71 s
[Task 17/25] Current/Best: 12.30/ 22.54 GFLOPS | Progress: (12/20) | 9.50 s
[Task 17/25] Current/Best: 18.04/ 22.54 GFLOPS | Progress: (16/20) | 11.91 s
[Task 17/25] Current/Best: 10.60/ 22.54 GFLOPS | Progress: (20/20) | 15.10 s Done.
+
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 16.76/ 16.76 GFLOPS | Progress: (4/20) | 4.80 s
[Task 18/25] Current/Best: 14.15/ 16.76 GFLOPS | Progress: (8/20) | 12.54 s
[Task 18/25] Current/Best: 17.28/ 17.28 GFLOPS | Progress: (12/20) | 16.90 s
[Task 18/25] Current/Best: 8.09/ 18.67 GFLOPS | Progress: (16/20) | 22.45 s
[Task 18/25] Current/Best: 20.55/ 20.55 GFLOPS | Progress: (20/20) | 30.38 s Done.
+
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 1.55/ 22.26 GFLOPS | Progress: (4/20) | 8.93 s
[Task 19/25] Current/Best: 3.09/ 22.26 GFLOPS | Progress: (8/20) | 13.10 s
[Task 19/25] Current/Best: 5.04/ 22.26 GFLOPS | Progress: (12/20) | 16.49 s
[Task 19/25] Current/Best: 15.57/ 22.26 GFLOPS | Progress: (16/20) | 19.00 s
[Task 19/25] Current/Best: 18.96/ 22.26 GFLOPS | Progress: (20/20) | 24.20 s Done.
+
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 10.52/ 10.52 GFLOPS | Progress: (4/20) | 4.38 s Done.
Done.
-
[Task 22/25] Current/Best: 8.88/ 18.72 GFLOPS | Progress: (12/20) | 12.07 s
[Task 22/25] Current/Best: 11.28/ 18.72 GFLOPS | Progress: (16/20) | 14.38 s
[Task 22/25] Current/Best: 14.86/ 19.06 GFLOPS | Progress: (20/20) | 15.95 s Done.
-
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 2.70/ 10.01 GFLOPS | Progress: (4/20) | 6.27 s
[Task 23/25] Current/Best: 12.19/ 12.19 GFLOPS | Progress: (8/20) | 10.30 s
[Task 23/25] Current/Best: 18.66/ 23.30 GFLOPS | Progress: (12/20) | 13.83 s
[Task 23/25] Current/Best: 10.67/ 23.30 GFLOPS | Progress: (16/20) | 20.73 s
[Task 23/25] Current/Best: 8.75/ 23.30 GFLOPS | Progress: (20/20) | 24.48 s Done.
-
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 5.12/ 5.51 GFLOPS | Progress: (4/20) | 9.12 s
[Task 24/25] Current/Best: 6.16/ 6.16 GFLOPS | Progress: (8/20) | 19.79 s
[Task 24/25] Current/Best: 4.66/ 6.16 GFLOPS | Progress: (12/20) | 30.76 s
[Task 24/25] Current/Best: 1.55/ 6.16 GFLOPS | Progress: (16/20) | 42.49 s
[Task 24/25] Current/Best: 8.25/ 8.32 GFLOPS | Progress: (20/20) | 53.46 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
[Task 25/25] Current/Best: 9.64/ 9.64 GFLOPS | Progress: (4/20) | 12.70 s
[Task 25/25] Current/Best: 3.66/ 9.64 GFLOPS | Progress: (8/20) | 23.63 s
[Task 25/25] Current/Best: 7.05/ 9.64 GFLOPS | Progress: (12/20) | 26.07 s
[Task 25/25] Current/Best: 6.34/ 9.74 GFLOPS | Progress: (16/20) | 36.71 s
[Task 25/25] Current/Best: 8.98/ 9.74 GFLOPS | Progress: (20/20) | 48.26 s
+
[Task 20/25] Current/Best: 1.57/ 10.52 GFLOPS | Progress: (8/20) | 7.98 s
[Task 20/25] Current/Best: 14.06/ 14.06 GFLOPS | Progress: (12/20) | 11.38 s
[Task 20/25] Current/Best: 15.66/ 20.57 GFLOPS | Progress: (16/20) | 15.27 s
[Task 20/25] Current/Best: 4.24/ 20.57 GFLOPS | Progress: (20/20) | 18.05 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 16.99/ 19.97 GFLOPS | Progress: (4/20) | 3.23 s
[Task 21/25] Current/Best: 10.75/ 19.97 GFLOPS | Progress: (8/20) | 5.27 s
[Task 21/25] Current/Best: 16.68/ 19.97 GFLOPS | Progress: (12/20) | 8.86 s
[Task 21/25] Current/Best: 8.66/ 19.97 GFLOPS | Progress: (16/20) | 12.06 s
[Task 21/25] Current/Best: 19.06/ 19.97 GFLOPS | Progress: (20/20) | 14.59 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 10.19/ 14.35 GFLOPS | Progress: (4/20
) | 6.67 s
[Task 22/25] Current/Best: 10.56/ 14.35 GFLOPS | Progress: (8/20) | 9.66 s
[Task 22/25] Current/Best: 5.31/ 16.13 GFLOPS | Progress: (12/20) | 12.20 s Done.
+
[Task 22/25] Current/Best: 18.00/ 18.54 GFLOPS | Progress: (16/20) | 13.92 s
[Task 22/25] Current/Best: 14.52/ 22.98 GFLOPS | Progress: (20/20) | 15.45 s Done.
+
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 2.68/ 23.67 GFLOPS | Progress: (4/20) | 6.10 s
[Task 23/25] Current/Best: 8.43/ 23.67 GFLOPS | Progress: (8/20) | 10.01 s
[Task 23/25] Current/Best: 9.90/ 23.67 GFLOPS | Progress: (12/20) | 13.49 s
[Task 23/25] Current/Best: 12.20/ 23.67 GFLOPS | Progress: (16/20) | 16.32 s
[Task 23/25] Current/Best: 9.05/ 24.01 GFLOPS | Progress: (20/20) | 18.52 s Done.
+
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 5.75/ 6.28 GFLOPS | Progress: (4/20) | 12.69 s
[Task 24/25] Current/Best: 3.79/ 6.28 GFLOPS | Progress: (8/20) | 23.62 s
[Task 24/25] Current/Best: 3.40/ 8.56 GFLOPS | Progress: (12/20) | 34.57 s
[Task 24/25] Current/Best: 3.23/ 8.56 GFLOPS | Progress: (16/20) | 41.17 s
[Task 24/25] Current/Best: 8.01/ 8.56 GFLOPS | Progress: (20/20) | 51.51 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+
[Task 25/25] Current/Best: 8.31/ 9.36 GFLOPS | Progress: (4/20) | 5.93 s
[Task 25/25] Current/Best: 5.12/ 9.42 GFLOPS | Progress: (8/20) | 16.84 s
[Task 25/25] Current/Best: 1.55/ 9.42 GFLOPS | Progress: (12/20) | 27.80 s
[Task 25/25] Current/Best: 1.55/ 9.42 GFLOPS | Progress: (16/20) | 33.45 s
[Task 25/25] Current/Best: 5.84/ 9.42 GFLOPS | Progress: (20/20) | 44.39 s
@@ -665,8 +664,8 @@ Verify that the optimized model runs and produces the same results:
.. code-block:: none
- class='n02123045 tabby, tabby cat' with probability=0.621102
- class='n02123159 tiger cat' with probability=0.356379
+ class='n02123045 tabby, tabby cat' with probability=0.621104
+ class='n02123159 tiger cat' with probability=0.356378
class='n02124075 Egyptian cat' with probability=0.019712
class='n02129604 tiger, Panthera tigris' with probability=0.001215
class='n04040759 radiator' with probability=0.000262
@@ -723,8 +722,8 @@ improvement in comparing the optimized model to the unoptimized model.
.. code-block:: none
- optimized: {'mean': 401.42029426000363, 'median': 400.3589346500007, 'std': 4.107752682740695}
- unoptimized: {'mean': 514.7913624499994, 'median': 514.5572220499957, 'std': 1.832343611061247}
+ optimized: {'mean': 410.78117241999735, 'median': 409.68679100000145, 'std': 2.8270131130514353}
+ unoptimized: {'mean': 524.7727405299996, 'median': 524.005764550003, 'std': 2.8968953730024656}
@@ -747,7 +746,7 @@ profiling/benchmarking.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 12 minutes 30.186 seconds)
+ **Total running time of the script:** ( 12 minutes 18.008 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 f313c03fb4..42f8718089 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
- 1.359e-07 secs/op
+ 1.256e-07 secs/op
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index 4203267123..f4d6f5bad5 100644
--- a/docs/_sources/tutorial/intro_topi.rst.txt
+++ b/docs/_sources/tutorial/intro_topi.rst.txt
@@ -277,7 +277,7 @@ As you can see, scheduled stages of computation have been accumulated and we can
.. code-block:: none
- [stage(a, placeholder(a, 0x594c300)), stage(b, placeholder(b, 0xe540250)), stage(T_add, compute(T_add, body=[a[ax0, ax1, ax2] + b[ax1, ax2]], axis=[T.iter_var(ax0, T.Range(0, 100), "DataPar", ""), T.iter_var(ax1, T.Range(0, 10), "DataPar", ""), T.iter_var(ax2, T.Range(0, 10), "DataPar", "")], reduce_axis=[], tag=broadcast, attrs={})), stage(T_multiply, compute(T_multiply, body=[a[ax0, ax1, ax2] * b[ax1, ax2]], axis=[T.iter_var(ax0, T.Range(0, 100), "DataPar", ""), T.iter_var(ax1, T.R [...]
+ [stage(a, placeholder(a, 0x22d93640)), stage(b, placeholder(b, 0x276de770)), 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 31642b60f7..c28f279916 100644
--- a/docs/_sources/tutorial/sg_execution_times.rst.txt
+++ b/docs/_sources/tutorial/sg_execution_times.rst.txt
@@ -5,32 +5,32 @@
Computation times
=================
-**15:59.338** total execution time for **tutorial** files:
+**15:37.302** total execution time for **tutorial** files:
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 12:30.186 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 12:18.008 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:32.780 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:21.768 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 00:57.147 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 00:58.973 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:36.083 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:35.943 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:20.573 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:20.244 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:01.548 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:01.361 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.850 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.846 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.169 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.159 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_uma.py` (``uma.py``) | 00:00.000 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``) | 00:00.000 | 0.0 MB |
-+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``) | 00:00.000 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_tvmc_python.py` (``tvmc_python.py``) | 00:00.000 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
+| :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``) | 00:00.000 | 0.0 MB |
++------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_install.py` (``install.py``) | 00:00.000 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
index 55f2adb048..4c35557c12 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -286,7 +286,7 @@ helper function to run a profile of the TVM generated code.
.. code-block:: none
Numpy running time: 0.000007
- naive: 0.000007
+ naive: 0.000008
@@ -392,7 +392,7 @@ compile and run this new schedule with the parallel operation applied:
.. code-block:: none
- parallel: 0.000007
+ parallel: 0.000006
@@ -504,10 +504,10 @@ We can now compare the different schedules
.. code-block:: none
Operator Timing Performance
- numpy 6.894450000345387e-06 1.0
- naive 6.664299999999999e-06 0.9666180768104986
- parallel 6.900100000000001e-06 1.0008194996924094
- vector 2.46599e-05 3.57677552216125
+ numpy 7.382000001143752e-06 1.0
+ naive 8.083799999999999e-06 1.0950690867986335
+ parallel 6.0816e-06 0.8238417771684814
+ vector 2.46195e-05 3.335071795744447
@@ -928,7 +928,7 @@ matrix multiplication.
.. code-block:: none
- Numpy running time: 0.018322
+ Numpy running time: 0.019867
@@ -986,7 +986,7 @@ optimizations.
.. code-block:: none
- none: 3.101824
+ none: 3.239250
@@ -1086,7 +1086,7 @@ schedule.
.. code-block:: none
- blocking: 0.292941
+ blocking: 0.297172
@@ -1170,7 +1170,7 @@ already cache friendly from our previous optimizations.
.. code-block:: none
- vectorization: 0.329768
+ vectorization: 0.335437
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1236,7 +1236,7 @@ more cache friendly.
.. code-block:: none
- loop permutation: 0.114206
+ loop permutation: 0.115665
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1327,7 +1327,7 @@ optimized schedule.
.. code-block:: none
- array packing: 0.109204
+ array packing: 0.109287
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1410,7 +1410,7 @@ to `C` when all the block results are ready.
.. code-block:: none
- block caching: 0.110825
+ block caching: 0.115502
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1484,7 +1484,7 @@ of thread-level parallelization.
.. code-block:: none
- parallelization: 0.146088
+ parallelization: 0.150991
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1554,13 +1554,13 @@ working, we can compare the results.
.. code-block:: none
Operator Timing Performance
- none 3.1018235676 1.0
- blocking 0.2929407008 0.09444144530330573
- vectorization 0.3297678629 0.10631419089872803
- loop permutation 0.1142055053 0.036818826993556306
- array packing 0.10920414780000001 0.035206434350647305
- block caching 0.11082455490000001 0.03572883901509241
- parallelization 0.1460879761 0.04709744861892125
+ none 3.2392502151 1.0
+ blocking 0.29717227910000005 0.09174106949648714
+ vectorization 0.33543742940000004 0.10355403476901355
+ loop permutation 0.11566527659999999 0.0357074226809704
+ array packing 0.1092867175 0.03373827591044126
+ block caching 0.1155023809 0.03565713459293057
+ parallelization 0.1509908888 0.04661291310444158
diff --git a/docs/commit_hash b/docs/commit_hash
index d1515e5e47..ced79c2965 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-c36ae1cb3a7a7a7bb843c48662ced26097650f90
+2e30e772a7395afa6674235e22943c433a1a3f32
diff --git a/docs/how_to/compile_models/from_darknet.html b/docs/how_to/compile_models/from_darknet.html
index ada48e97ec..83a3378c58 100644
--- a/docs/how_to/compile_models/from_darknet.html
+++ b/docs/how_to/compile_models/from_darknet.html
@@ -585,7 +585,7 @@ class:['truck 0.9266'] left:471 top:83 right:689 bottom:169
class:['bicycle 0.9984'] left:111 top:113 right:577 bottom:447
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 19.680 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 19.010 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-compile-models-from-darknet-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/7716f96385bd5abb6e822041e285be54/from_darknet.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">from_darknet.py</span></code></a></p>
diff --git a/docs/how_to/compile_models/from_keras.html b/docs/how_to/compile_models/from_keras.html
index 5768791fc9..2f7121096f 100644
--- a/docs/how_to/compile_models/from_keras.html
+++ b/docs/how_to/compile_models/from_keras.html
@@ -506,7 +506,7 @@ Tensorflow is also required since it’s used as the default backend of keras.</
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Relay top-1 id: 285, class name: Egyptian cat
1/1 [==============================] - ETA: 0s
-1/1 [==============================] - 1s 928ms/step
+1/1 [==============================] - 1s 937ms/step
Keras top-1 id: 285, class name: Egyptian cat
</pre></div>
</div>
diff --git a/docs/how_to/compile_models/from_mxnet.html b/docs/how_to/compile_models/from_mxnet.html
index 7e998c6f71..b2f1abf663 100644
--- a/docs/how_to/compile_models/from_mxnet.html
+++ b/docs/how_to/compile_models/from_mxnet.html
@@ -439,7 +439,7 @@
<span class="nb">print</span><span class="p">(</span><span class="s2">"x"</span><span class="p">,</span> <a href="https://docs.python.org/3/library/stdtypes.html#tuple" title="builtins.tuple" class="sphx-glr-backref-module-builtins sphx-glr-backref-type-py-class sphx-glr-backref-instance"><span class="n">x</span><span class="o">.</span><span class="n">shape</span></a><span class="p">)</span>
</pre></div>
</div>
-<img src="../../_images/sphx_glr_from_mxnet_001.png" srcset="../../_images/sphx_glr_from_mxnet_001.png" alt="from mxnet" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip02131915-cf90-43bd-a4e4-e8fdd67e5c71 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.zip533579d2-278b-4bde-84d7-3b4a17fc931c 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 7b0cc95c6f..ce124b57bd 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -449,14 +449,14 @@ 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 | 7.99M/41.5M [00:00<00:00, 39.4MB/s]
- 35%|###4 | 14.3M/41.5M [00:00<00:00, 43.6MB/s]
- 47%|####6 | 19.3M/41.5M [00:00<00:00, 46.5MB/s]
- 58%|#####7 | 23.9M/41.5M [00:00<00:00, 45.9MB/s]
- 68%|######8 | 28.4M/41.5M [00:00<00:00, 45.5MB/s]
- 79%|#######8 | 32.8M/41.5M [00:00<00:00, 38.9MB/s]
- 92%|#########2| 38.3M/41.5M [00:00<00:00, 40.3MB/s]
-100%|##########| 41.5M/41.5M [00:01<00:00, 42.0MB/s]
+ 19%|#9 | 7.99M/41.5M [00:00<00:00, 49.0MB/s]
+ 35%|###4 | 14.3M/41.5M [00:00<00:00, 39.8MB/s]
+ 44%|####3 | 18.2M/41.5M [00:00<00:00, 38.3MB/s]
+ 54%|#####3 | 22.3M/41.5M [00:00<00:00, 39.1MB/s]
+ 63%|######2 | 26.1M/41.5M [00:00<00:00, 36.9MB/s]
+ 77%|#######7 | 32.0M/41.5M [00:00<00:00, 42.7MB/s]
+ 92%|#########2| 38.3M/41.5M [00:01<00:00, 30.5MB/s]
+100%|##########| 41.5M/41.5M [00:01<00:00, 34.6MB/s]
</pre></div>
</div>
</div>
diff --git a/docs/how_to/compile_models/from_pytorch.html b/docs/how_to/compile_models/from_pytorch.html
index f597bed239..a9c99eeda7 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -432,13 +432,12 @@ be unstable.</p>
Downloading: "https://download.pytorch.org/models/resnet18-f37072fd.pth" to /workspace/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
0%| | 0.00/44.7M [00:00<?, ?B/s]
- 18%|#7 | 7.99M/44.7M [00:00<00:00, 49.1MB/s]
- 36%|###5 | 16.0M/44.7M [00:00<00:00, 60.1MB/s]
- 54%|#####4 | 24.2M/44.7M [00:00<00:00, 69.7MB/s]
- 70%|######9 | 31.1M/44.7M [00:00<00:00, 51.5MB/s]
- 82%|########2 | 36.6M/44.7M [00:00<00:00, 46.7MB/s]
- 93%|#########2| 41.5M/44.7M [00:00<00:00, 43.4MB/s]
-100%|##########| 44.7M/44.7M [00:00<00:00, 49.1MB/s]
+ 14%|#4 | 6.30M/44.7M [00:00<00:00, 65.5MB/s]
+ 34%|###3 | 15.2M/44.7M [00:00<00:00, 81.7MB/s]
+ 54%|#####3 | 24.0M/44.7M [00:00<00:00, 78.4MB/s]
+ 72%|#######1 | 32.1M/44.7M [00:00<00:00, 79.4MB/s]
+ 90%|########9 | 40.0M/44.7M [00:00<00:00, 65.1MB/s]
+100%|##########| 44.7M/44.7M [00:00<00:00, 60.6MB/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 68713f3650..595daa44d2 100644
--- a/docs/how_to/compile_models/from_tensorflow.html
+++ b/docs/how_to/compile_models/from_tensorflow.html
@@ -649,7 +649,7 @@ banana (score = 0.00022)
desk (score = 0.00019)
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 23.311 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 22.434 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 4be487b850..4fa8e2c48b 100644
--- a/docs/how_to/compile_models/sg_execution_times.html
+++ b/docs/how_to/compile_models/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-compile-models-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>06:34.450</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>06:31.248</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 81%" />
@@ -349,43 +349,43 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="from_tensorflow.html#sphx-glr-how-to-compile-models-from-tensorflow-py"><span class="std std-ref">Compile Tensorflow Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tensorflow.py</span></code>)</p></td>
-<td><p>01:23.311</p></td>
+<td><p>01:22.434</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>01:19.680</p></td>
+<td><p>01:19.010</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="from_paddle.html#sphx-glr-how-to-compile-models-from-paddle-py"><span class="std std-ref">Compile PaddlePaddle Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_paddle.py</span></code>)</p></td>
-<td><p>00:53.899</p></td>
+<td><p>00:53.764</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>00:36.783</p></td>
+<td><p>00:36.826</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>00:31.246</p></td>
+<tr class="row-odd"><td><p><a class="reference internal" href="from_mxnet.html#sphx-glr-how-to-compile-models-from-mxnet-py"><span class="std std-ref">Compile MXNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_mxnet.py</span></code>)</p></td>
+<td><p>00:31.177</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>00:30.816</p></td>
+<tr class="row-even"><td><p><a class="reference internal" href="from_coreml.html#sphx-glr-how-to-compile-models-from-coreml-py"><span class="std std-ref">Compile CoreML Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_coreml.py</span></code>)</p></td>
+<td><p>00:30.700</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>00:28.290</p></td>
+<td><p>00:27.622</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>00:25.268</p></td>
+<td><p>00:25.010</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>00:22.520</p></td>
+<td><p>00:22.047</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>00:02.637</p></td>
+<td><p>00:02.657</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 db06fe8298..7b9b588ca0 100644
--- a/docs/how_to/deploy_models/deploy_model_on_adreno.html
+++ b/docs/how_to/deploy_models/deploy_model_on_adreno.html
@@ -920,7 +920,7 @@ Top5 predictions:
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 2753.6120 2753.0844 2757.5755 2751.7726 1.6565
+ 2753.7866 2752.9268 2758.3577 2750.6551 2.1445
</pre></div>
</div>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-model-on-adreno-py">
diff --git a/docs/how_to/deploy_models/deploy_model_on_android.html b/docs/how_to/deploy_models/deploy_model_on_android.html
index 6ec42d0eab..4ac5a06768 100644
--- a/docs/how_to/deploy_models/deploy_model_on_android.html
+++ b/docs/how_to/deploy_models/deploy_model_on_android.html
@@ -662,7 +662,7 @@ to the remote android device.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 15.7103 15.5860 16.5707 15.4881 0.3229
+ 16.4693 16.3055 17.3688 15.9610 0.4562
</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 e1a30e6d30..0ce8e2a9df 100644
--- a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
+++ b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
@@ -454,34 +454,31 @@ 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<00:06, 24.5MB/s]
- 5%|5 | 8.65M/170M [00:00<00:06, 24.5MB/s]
- 8%|8 | 14.3M/170M [00:00<00:05, 29.0MB/s]
- 10%|# | 17.1M/170M [00:00<00:06, 26.0MB/s]
- 13%|#3 | 22.3M/170M [00:00<00:04, 31.6MB/s]
- 15%|#4 | 25.4M/170M [00:00<00:05, 27.6MB/s]
- 19%|#8 | 32.0M/170M [00:01<00:04, 32.6MB/s]
- 24%|##3 | 40.0M/170M [00:01<00:03, 36.5MB/s]
- 28%|##8 | 48.0M/170M [00:01<00:03, 42.1MB/s]
- 31%|### | 52.0M/170M [00:01<00:03, 36.8MB/s]
- 33%|###2 | 56.0M/170M [00:01<00:03, 36.6MB/s]
- 37%|###6 | 62.3M/170M [00:01<00:03, 35.3MB/s]
- 39%|###8 | 65.7M/170M [00:02<00:03, 33.5MB/s]
- 42%|####2 | 72.0M/170M [00:02<00:02, 38.7MB/s]
- 47%|####7 | 80.0M/170M [00:02<00:01, 47.2MB/s]
- 52%|#####1 | 88.0M/170M [00:02<00:01, 51.7MB/s]
- 56%|#####5 | 94.3M/170M [00:02<00:01, 52.0MB/s]
- 60%|###### | 102M/170M [00:02<00:01, 59.5MB/s]
- 65%|######4 | 110M/170M [00:02<00:00, 64.7MB/s]
- 68%|######8 | 116M/170M [00:02<00:00, 57.6MB/s]
- 72%|#######1 | 122M/170M [00:03<00:00, 55.3MB/s]
- 75%|#######5 | 128M/170M [00:03<00:00, 52.4MB/s]
- 80%|######## | 136M/170M [00:03<00:00, 50.2MB/s]
- 83%|########3 | 141M/170M [00:03<00:00, 50.5MB/s]
- 86%|########6 | 146M/170M [00:03<00:00, 42.7MB/s]
- 93%|#########3| 158M/170M [00:03<00:00, 61.5MB/s]
- 97%|#########7| 165M/170M [00:03<00:00, 54.6MB/s]
-100%|##########| 170M/170M [00:04<00:00, 43.1MB/s]
+ 4%|3 | 6.30M/170M [00:00<00:04, 40.9MB/s]
+ 6%|6 | 10.2M/170M [00:00<00:04, 35.6MB/s]
+ 9%|9 | 16.0M/170M [00:00<00:04, 34.0MB/s]
+ 13%|#3 | 22.3M/170M [00:00<00:03, 42.2MB/s]
+ 18%|#7 | 30.3M/170M [00:00<00:02, 49.5MB/s]
+ 21%|## | 35.2M/170M [00:00<00:03, 43.5MB/s]
+ 27%|##7 | 46.3M/170M [00:00<00:02, 61.4MB/s]
+ 31%|###1 | 52.7M/170M [00:01<00:02, 49.2MB/s]
+ 34%|###4 | 58.1M/170M [00:01<00:02, 48.3MB/s]
+ 38%|###7 | 64.0M/170M [00:01<00:02, 48.6MB/s]
+ 42%|####2 | 72.0M/170M [00:01<00:01, 51.6MB/s]
+ 47%|####7 | 80.0M/170M [00:01<00:01, 53.4MB/s]
+ 52%|#####1 | 88.0M/170M [00:01<00:01, 57.0MB/s]
+ 56%|#####5 | 94.3M/170M [00:02<00:01, 52.5MB/s]
+ 60%|###### | 102M/170M [00:02<00:01, 58.8MB/s]
+ 64%|######3 | 108M/170M [00:02<00:01, 56.6MB/s]
+ 68%|######8 | 116M/170M [00:02<00:00, 63.7MB/s]
+ 72%|#######2 | 123M/170M [00:02<00:00, 56.1MB/s]
+ 75%|#######5 | 128M/170M [00:02<00:00, 54.7MB/s]
+ 80%|######## | 136M/170M [00:02<00:00, 62.0MB/s]
+ 85%|########4 | 144M/170M [00:02<00:00, 53.5MB/s]
+ 88%|########8 | 150M/170M [00:03<00:00, 51.4MB/s]
+ 92%|#########1| 155M/170M [00:03<00:00, 48.0MB/s]
+ 97%|#########7| 165M/170M [00:03<00:00, 54.1MB/s]
+100%|##########| 170M/170M [00:03<00:00, 50.9MB/s]
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/nn/functional.py:3897: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
for i in range(dim)
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/detection/anchor_utils.py:124: UserWarning: __floordiv__ is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode=& [...]
@@ -579,7 +576,7 @@ 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 28.687 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 27.823 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 9e439b0398..b71a426044 100644
--- a/docs/how_to/deploy_models/deploy_prequantized.html
+++ b/docs/how_to/deploy_models/deploy_prequantized.html
@@ -495,9 +495,8 @@ training. Other models require a full post training calibration.</p>
Downloading: "https://download.pytorch.org/models/mobilenet_v2-b0353104.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2-b0353104.pth
0%| | 0.00/13.6M [00:00<?, ?B/s]
- 59%|#####8 | 7.99M/13.6M [00:00<00:00, 57.9MB/s]
-100%|#########9| 13.5M/13.6M [00:00<00:00, 45.4MB/s]
-100%|##########| 13.6M/13.6M [00:00<00:00, 47.1MB/s]
+ 59%|#####8 | 7.99M/13.6M [00:00<00:00, 59.9MB/s]
+100%|##########| 13.6M/13.6M [00:00<00:00, 63.9MB/s]
</pre></div>
</div>
</div>
@@ -588,7 +587,7 @@ output values are identical out of 1000 outputs from mobilenet v2.</p>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 90.3866 90.3266 92.8951 90.0252 0.4161
+ 90.6068 90.5499 91.6871 90.2145 0.2587
</pre></div>
</div>
<div class="admonition note">
@@ -627,7 +626,7 @@ This includes support for the VNNI 8 bit dot product instruction (CascadeLake or
<div class="section" id="deploy-a-quantized-tflite-model">
<h2>Deploy a quantized TFLite Model<a class="headerlink" href="#deploy-a-quantized-tflite-model" title="Permalink to this headline">¶</a></h2>
<p>TODO</p>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 14.058 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 13.472 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 db5b581b24..a6d36685e2 100644
--- a/docs/how_to/deploy_models/deploy_prequantized_tflite.html
+++ b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
@@ -580,7 +580,7 @@ TFLite Top-5 labels: [387 102 386 341 349]
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 120.4425 120.4263 124.4521 119.7056 0.4887
+ 119.3548 119.3192 123.5348 117.3429 0.7662
</pre></div>
</div>
<div class="admonition note">
@@ -608,7 +608,7 @@ network for ARM CPU</span></a>.</p></li>
</ul>
</div></blockquote>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 41.952 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 32.431 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 9aeb7889ad..2ce3be9b37 100644
--- a/docs/how_to/deploy_models/deploy_quantized.html
+++ b/docs/how_to/deploy_models/deploy_quantized.html
@@ -521,7 +521,7 @@ for calibration. But the accuracy might be impacted.</p>
DeprecationWarning,
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 50.624 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 50.647 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 2faae496f0..0cf3a9662a 100644
--- a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
+++ b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
@@ -463,24 +463,23 @@ to your device.</p>
Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/ssd_512_resnet50_v1_voc-9c8b225a.zip...
0%| | 0/132723 [00:00<?, ?KB/s]
- 3%|2 | 3379/132723 [00:00<00:03, 33787.94KB/s]
- 8%|7 | 10298/132723 [00:00<00:02, 54608.05KB/s]
- 14%|#3 | 18379/132723 [00:00<00:01, 66569.98KB/s]
- 20%|## | 26553/132723 [00:00<00:01, 72555.86KB/s]
- 26%|##6 | 34718/132723 [00:00<00:01, 75824.82KB/s]
- 32%|###2 | 42862/132723 [00:00<00:01, 77732.51KB/s]
- 38%|###8 | 50928/132723 [00:00<00:01, 78686.63KB/s]
- 45%|####4 | 59078/132723 [00:00<00:00, 79580.64KB/s]
- 51%|##### | 67208/132723 [00:00<00:00, 80116.39KB/s]
- 57%|#####6 | 75394/132723 [00:01<00:00, 80652.35KB/s]
- 63%|######2 | 83533/132723 [00:01<00:00, 80876.29KB/s]
- 69%|######9 | 91640/132723 [00:01<00:00, 80932.71KB/s]
- 75%|#######5 | 99762/132723 [00:01<00:00, 81017.01KB/s]
- 81%|########1 | 107929/132723 [00:01<00:00, 81210.81KB/s]
- 87%|########7 | 116075/132723 [00:01<00:00, 81282.84KB/s]
- 94%|#########3| 124220/132723 [00:01<00:00, 81331.19KB/s]
-100%|#########9| 132534/132723 [00:01<00:00, 81873.09KB/s]
-100%|##########| 132723/132723 [00:01<00:00, 77858.45KB/s]
+ 2%|1 | 2605/132723 [00:00<00:04, 26045.86KB/s]
+ 6%|6 | 8383/132723 [00:00<00:02, 44657.14KB/s]
+ 12%|#2 | 16190/132723 [00:00<00:01, 59901.02KB/s]
+ 19%|#8 | 24805/132723 [00:00<00:01, 70256.82KB/s]
+ 25%|##4 | 32879/132723 [00:00<00:01, 74030.58KB/s]
+ 31%|###1 | 41471/132723 [00:00<00:01, 78068.40KB/s]
+ 38%|###7 | 50022/132723 [00:00<00:01, 80496.63KB/s]
+ 44%|####4 | 58660/132723 [00:00<00:00, 82367.69KB/s]
+ 51%|##### | 67243/132723 [00:00<00:00, 83446.90KB/s]
+ 57%|#####7 | 75896/132723 [00:01<00:00, 84397.08KB/s]
+ 64%|######3 | 84525/132723 [00:01<00:00, 84973.34KB/s]
+ 70%|####### | 93215/132723 [00:01<00:00, 85555.58KB/s]
+ 77%|#######6 | 101849/132723 [00:01<00:00, 85790.86KB/s]
+ 83%|########3 | 110556/132723 [00:01<00:00, 86171.44KB/s]
+ 90%|########9 | 119208/132723 [00:01<00:00, 86274.60KB/s]
+ 96%|#########6| 127911/132723 [00:01<00:00, 86500.86KB/s]
+100%|##########| 132723/132723 [00:01<00:00, 80060.59KB/s]
</pre></div>
</div>
<p>Create TVM runtime and do inference
@@ -519,7 +518,7 @@ Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from h
<span class="n">plt</span><span class="o">.</span><span class="n">show</span><span class="p">()</span>
</pre></div>
</div>
-<img src="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" srcset="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" alt="deploy ssd gluoncv" class = "sphx-glr-single-img"/><p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 43.106 seconds)</p>
+<img src="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" srcset="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" alt="deploy ssd gluoncv" class = "sphx-glr-single-img"/><p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 42.134 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 8c034aec62..24b8dc91f0 100644
--- a/docs/how_to/deploy_models/sg_execution_times.html
+++ b/docs/how_to/deploy_models/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-deploy-models-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>15:29.939</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>15:17.691</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 86%" />
@@ -349,39 +349,39 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="deploy_ssd_gluoncv.html#sphx-glr-how-to-deploy-models-deploy-ssd-gluoncv-py"><span class="std std-ref">Deploy Single Shot Multibox Detector(SSD) model</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_ssd_gluoncv.py</span></code>)</p></td>
-<td><p>03:43.106</p></td>
+<td><p>03:42.134</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>03:28.687</p></td>
+<td><p>03:27.823</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>02:41.952</p></td>
+<td><p>02:32.431</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="deploy_quantized.html#sphx-glr-how-to-deploy-models-deploy-quantized-py"><span class="std std-ref">Deploy a Quantized Model on Cuda</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_quantized.py</span></code>)</p></td>
-<td><p>01:50.624</p></td>
+<td><p>01:50.647</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>01:14.058</p></td>
+<td><p>01:13.472</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="deploy_model_on_adreno.html#sphx-glr-how-to-deploy-models-deploy-model-on-adreno-py"><span class="std std-ref">Deploy the Pretrained Model on Adreno</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_adreno.py</span></code>)</p></td>
-<td><p>00:56.124</p></td>
+<td><p>00:56.069</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="deploy_model_on_android.html#sphx-glr-how-to-deploy-models-deploy-model-on-android-py"><span class="std std-ref">Deploy the Pretrained Model on Android</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_android.py</span></code>)</p></td>
-<td><p>00:40.685</p></td>
+<td><p>00:40.441</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="deploy_model_on_nano.html#sphx-glr-how-to-deploy-models-deploy-model-on-nano-py"><span class="std std-ref">Deploy the Pretrained Model on Jetson Nano</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_nano.py</span></code>)</p></td>
-<td><p>00:27.490</p></td>
+<td><p>00:27.542</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="deploy_model_on_rasp.html#sphx-glr-how-to-deploy-models-deploy-model-on-rasp-py"><span class="std std-ref">Deploy the Pretrained Model on Raspberry Pi</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_rasp.py</span></code>)</p></td>
-<td><p>00:27.207</p></td>
+<td><p>00:27.127</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="deploy_sparse.html#sphx-glr-how-to-deploy-models-deploy-sparse-py"><span class="std std-ref">Deploy a Hugging Face Pruned Model on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_sparse.py</span></code>)</p></td>
diff --git a/docs/how_to/extend_tvm/bring_your_own_datatypes.html b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
index b324656b0e..ceaa75ad66 100644
--- a/docs/how_to/extend_tvm/bring_your_own_datatypes.html
+++ b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
@@ -619,7 +619,7 @@ In this alpha state of the Bring Your Own Datatypes framework, we have not imple
<span class="n">module</span><span class="p">,</span> <a href="https://docs.python.org/3/library/stdtypes.html#dict" title="builtins.dict" class="sphx-glr-backref-module-builtins sphx-glr-backref-type-py-class sphx-glr-backref-instance"><span class="n">params</span></a> <span class="o">=</span> <span class="n">get_mobilenet</span><span class="p">()</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip29b949b8-121d-4be9-864b-50e9878edb73 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.zipc9c140f0-8012-4216-bf1a-486ce56b5235 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 816ea87f08..1d81672d33 100644
--- a/docs/how_to/extend_tvm/sg_execution_times.html
+++ b/docs/how_to/extend_tvm/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-extend-tvm-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:53.009</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:52.742</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -349,15 +349,15 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="bring_your_own_datatypes.html#sphx-glr-how-to-extend-tvm-bring-your-own-datatypes-py"><span class="std std-ref">Bring Your Own Datatypes to TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">bring_your_own_datatypes.py</span></code>)</p></td>
-<td><p>00:49.188</p></td>
+<td><p>00:48.972</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="use_pass_instrument.html#sphx-glr-how-to-extend-tvm-use-pass-instrument-py"><span class="std std-ref">How to Use TVM Pass Instrument</span></a> (<code class="docutils literal notranslate"><span class="pre">use_pass_instrument.py</span></code>)</p></td>
-<td><p>00:02.733</p></td>
+<td><p>00:02.697</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>00:01.079</p></td>
+<td><p>00:01.064</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 e3ddc13160..6794a553a8 100644
--- a/docs/how_to/extend_tvm/use_pass_instrument.html
+++ b/docs/how_to/extend_tvm/use_pass_instrument.html
@@ -526,10 +526,10 @@ profile the execution time of each passes.</p>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Printing results of timing profile...
-InferType: 22161us [22161us] (48.95%; 48.95%)
-FoldScaleAxis: 23109us [7us] (51.05%; 51.05%)
- FoldConstant: 23102us [1705us] (51.03%; 99.97%)
- InferType: 21397us [21397us] (47.26%; 92.62%)
+InferType: 22397us [22397us] (48.70%; 48.70%)
+FoldScaleAxis: 23591us [7us] (51.30%; 51.30%)
+ FoldConstant: 23583us [1737us] (51.28%; 99.97%)
+ InferType: 21847us [21847us] (47.50%; 92.64%)
</pre></div>
</div>
</div>
@@ -551,10 +551,10 @@ Refer to following sections and <a class="reference internal" href="../../refere
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Printing results of timing profile...
-InferType: 21509us [21509us] (48.47%; 48.47%)
-FoldScaleAxis: 22864us [5us] (51.53%; 51.53%)
- FoldConstant: 22859us [1659us] (51.52%; 99.98%)
- InferType: 21200us [21200us] (47.78%; 92.74%)
+InferType: 21925us [21925us] (48.29%; 48.29%)
+FoldScaleAxis: 23478us [6us] (51.71%; 51.71%)
+ FoldConstant: 23473us [1741us] (51.70%; 99.98%)
+ InferType: 21732us [21732us] (47.86%; 92.58%)
</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 79518a01a2..7b1b0a5bbb 100644
--- a/docs/how_to/optimize_operators/opt_conv_cuda.html
+++ b/docs/how_to/optimize_operators/opt_conv_cuda.html
@@ -575,7 +575,7 @@ latency of convolution.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Convolution: </span><span class="si">%f</span><span class="s2"> ms"</span> <span class="o">%</span> <span class="p">(</span><span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">w</span><span class="p">,</span> <span class="n">b</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span> <span class="o">*</span> <span cl [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 54.091423 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 33.637153 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 cde1a0710d..3305143f3b 100644
--- a/docs/how_to/optimize_operators/opt_conv_tensorcore.html
+++ b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
@@ -867,7 +867,7 @@ be able to run on our build server</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"conv2d with tensor core: </span><span class="si">%f</span><span class="s2"> ms"</span> <span class="o">%</span> <span class="p">(</span><span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">w</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span> <span class="o">* [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 6.593321 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 13.335760 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 5350649bc9..8fe9e9d64d 100644
--- a/docs/how_to/optimize_operators/opt_gemm.html
+++ b/docs/how_to/optimize_operators/opt_gemm.html
@@ -472,8 +472,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
<span class="nb">print</span><span class="p">(</span><span class="s2">"Baseline: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018411
-Baseline: 3.234937
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018666
+Baseline: 3.252249
</pre></div>
</div>
<p>In TVM, we can always inspect lower level IR to debug or optimize our schedule.
@@ -532,7 +532,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt1: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.299118
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.301624
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -589,7 +589,7 @@ vastly.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt2: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.333509
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.335315
</pre></div>
</div>
<p>Here is the generated IR after vectorization.</p>
@@ -644,7 +644,7 @@ the access pattern for A matrix is more cache friendly.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt3: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.115660
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.114184
</pre></div>
</div>
<p>Here is the generated IR after loop permutation.</p>
@@ -721,7 +721,7 @@ flattening.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt4: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109347
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109062
</pre></div>
</div>
<p>Here is the generated IR after array packing.</p>
@@ -799,7 +799,7 @@ write to C when all the block results are ready.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt5: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.110709
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.114539
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -879,7 +879,7 @@ class Module:
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt6: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">opt6_time</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.146349
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.151849
</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 ce758f56ba..3086ece789 100644
--- a/docs/how_to/optimize_operators/sg_execution_times.html
+++ b/docs/how_to/optimize_operators/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-optimize-operators-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:34.251</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:34.787</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -349,15 +349,15 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="opt_gemm.html#sphx-glr-how-to-optimize-operators-opt-gemm-py"><span class="std std-ref">How to optimize GEMM on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_gemm.py</span></code>)</p></td>
-<td><p>00:31.679</p></td>
+<td><p>00:32.109</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>00:01.485</p></td>
+<td><p>00:01.598</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>00:01.087</p></td>
+<td><p>00:01.080</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 9bd0de6aad..e82efe6239 100644
--- a/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
+++ b/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-tune-with-autoscheduler-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>09:47.420</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>09:54.297</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 85%" />
@@ -349,27 +349,27 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="tune_conv2d_layer_cuda.html#sphx-glr-how-to-tune-with-autoscheduler-tune-conv2d-layer-cuda-py"><span class="std std-ref">Auto-scheduling a Convolution Layer for GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_conv2d_layer_cuda.py</span></code>)</p></td>
-<td><p>05:59.876</p></td>
+<td><p>06:08.809</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="tune_network_x86.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-x86-py"><span class="std std-ref">Auto-scheduling a Neural Network for x86 CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_x86.py</span></code>)</p></td>
-<td><p>01:40.136</p></td>
+<td><p>01:40.129</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>01:07.619</p></td>
+<td><p>01:07.165</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>00:31.899</p></td>
+<td><p>00:30.521</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>00:14.637</p></td>
+<td><p>00:14.574</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>00:13.252</p></td>
+<td><p>00:13.099</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 9e46956f68..3dcb8177aa 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html
@@ -506,181 +506,797 @@ class Module:
def main(data: T.Buffer((1, 512, 7, 7), "float32"), kernel: T.Buffer((512, 512, 3, 3), "float32"), bias: T.Buffer((1, 512, 1, 1), "float32"), compute: T.Buffer((1, 512, 7, 7), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
blockIdx_x = T.env_thread("blockIdx.x")
- T.launch_thread(blockIdx_x, 16)
- conv2d_nchw = T.allocate([2], "float32", "local")
- pad_temp_shared = T.allocate([648], "float32", "shared")
+ T.launch_thread(blockIdx_x, 56)
+ conv2d_nchw = T.allocate([16], "float32", "local")
+ pad_temp_shared = T.allocate([108], "float32", "shared")
kernel_shared = T.allocate([2304], "float32", "shared")
threadIdx_x = T.env_thread("threadIdx.x")
- T.launch_thread(threadIdx_x, 784)
- conv2d_nchw_1 = T.Buffer((2,), data=conv2d_nchw, scope="local", align=8)
+ T.launch_thread(threadIdx_x, 28)
+ conv2d_nchw_1 = T.Buffer((16,), data=conv2d_nchw, scope="local", align=16)
conv2d_nchw_1[0] = T.float32(0)
+ conv2d_nchw_1[4] = T.float32(0)
+ conv2d_nchw_1[8] = T.float32(0)
+ conv2d_nchw_1[12] = T.float32(0)
conv2d_nchw_1[1] = T.float32(0)
- for rc_outer_outer in range(64):
- cse_var_1: T.int32 = rc_outer_outer * 72
+ conv2d_nchw_1[5] = T.float32(0)
+ conv2d_nchw_1[9] = T.float32(0)
+ conv2d_nchw_1[13] = T.float32(0)
+ conv2d_nchw_1[2] = T.float32(0)
+ conv2d_nchw_1[6] = T.float32(0)
+ conv2d_nchw_1[10] = T.float32(0)
+ conv2d_nchw_1[14] = T.float32(0)
+ conv2d_nchw_1[3] = T.float32(0)
+ conv2d_nchw_1[7] = T.float32(0)
+ conv2d_nchw_1[11] = T.float32(0)
+ conv2d_nchw_1[15] = T.float32(0)
+ for rc_outer_outer in range(128):
+ cse_var_2: T.int32 = rc_outer_outer * 196
+ cse_var_1: T.int32 = rc_outer_outer * 36
threadIdx_x_1 = T.env_thread("threadIdx.x")
- pad_temp_shared_1 = T.Buffer((648,), data=pad_temp_shared, scope="shared")
- with T.launch_thread(threadIdx_x_1, 784):
- if T.likely(threadIdx_x_1 < 648):
- data_1 = T.Buffer((25088,), data=data.data)
- pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(9 <= threadIdx_x_1 % 81 and threadIdx_x_1 % 81 < 72 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[rc_outer_outer * 392 + threadIdx_x_1 // 81 * 49 + threadIdx_x_1 % 81 // 9 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
+ pad_temp_shared_1 = T.Buffer((108,), data=pad_temp_shared, scope="shared")
+ data_1 = T.Buffer((25088,), data=data.data)
+ with T.launch_thread(threadIdx_x_1, 28):
+ pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(1 <= threadIdx_x_1 % 27 // 9 + blockIdx_x % 7 and threadIdx_x_1 % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + threadIdx_x_1 // 27 * 49 + threadIdx_x_1 % 27 // 9 * 7 + blockIdx_x % 7 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 28):
+ pad_temp_shared_1[threadIdx_x_1 + 28] = T.if_then_else(1 <= (threadIdx_x_1 + 1) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 1) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 1) % 9 and (threadIdx_x_1 + 1) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 28) // 27 * 49 + (threadIdx_x_1 + 1) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 1) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 28):
+ pad_temp_shared_1[threadIdx_x_1 + 56] = T.if_then_else(1 <= (threadIdx_x_1 + 2) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 2) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 2) % 9 and (threadIdx_x_1 + 2) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 56) // 27 * 49 + (threadIdx_x_1 + 2) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 2) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 28):
+ if T.likely(threadIdx_x_1 < 24):
+ pad_temp_shared_1[threadIdx_x_1 + 84] = T.if_then_else(1 <= (threadIdx_x_1 + 3) // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 3) // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 3) % 9 and (threadIdx_x_1 + 3) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 84) // 27 * 49 + (threadIdx_x_1 + 3) // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 3) % 9 - 8], T.float32(0))
threadIdx_x_2 = T.env_thread("threadIdx.x")
kernel_shared_1 = T.Buffer((2304,), data=kernel_shared, scope="shared")
kernel_1 = T.Buffer((2359296,), data=kernel.data)
- with T.launch_thread(threadIdx_x_2, 784):
- kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x * 147456 + threadIdx_x_2 // 72 * 4608 + cse_var_1 + threadIdx_x_2 % 72]
- with T.launch_thread(threadIdx_x_2, 784):
- kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x * 147456 + (threadIdx_x_2 + 784) // 72 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 72 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 784):
- if T.likely(threadIdx_x_2 < 736):
- kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x * 147456 + (threadIdx_x_2 + 1568) // 72 * 4608 + cse_var_1 + (threadIdx_x_2 + 56) % 72 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 49 * 144]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 49 * 144 + 72]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 49 * 144 + 1]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 49 * 144 + 73]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 49 * 144 + 2]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 49 * 144 + 74]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 49 * 144 + 9]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 49 * 144 + 81]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 49 * 144 + 10]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 49 * 144 + 82]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 49 * 144 + 11]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 49 * 144 + 83]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 162] * kernel_shared_1[threadIdx_x // 49 * 144 + 18]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 162] * kernel_shared_1[threadIdx_x // 49 * 144 + 90]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 163] * kernel_shared_1[threadIdx_x // 49 * 144 + 19]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 163] * kernel_shared_1[threadIdx_x // 49 * 144 + 91]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 164] * kernel_shared_1[threadIdx_x // 49 * 144 + 20]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 164] * kernel_shared_1[threadIdx_x // 49 * 144 + 92]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 243] * kernel_shared_1[threadIdx_x // 49 * 144 + 27]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 243] * kernel_shared_1[threadIdx_x // 49 * 144 + 99]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 244] * kernel_shared_1[threadIdx_x // 49 * 144 + 28]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 244] * kernel_shared_1[threadIdx_x // 49 * 144 + 100]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 245] * kernel_shared_1[threadIdx_x // 49 * 144 + 29]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 245] * kernel_shared_1[threadIdx_x // 49 * 144 + 101]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 324] * kernel_shared_1[threadIdx_x // 49 * 144 + 36]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 324] * kernel_shared_1[threadIdx_x // 49 * 144 + 108]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 325] * kernel_shared_1[threadIdx_x // 49 * 144 + 37]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 325] * kernel_shared_1[threadIdx_x // 49 * 144 + 109]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 326] * kernel_shared_1[threadIdx_x // 49 * 144 + 38]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 326] * kernel_shared_1[threadIdx_x // 49 * 144 + 110]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 405] * kernel_shared_1[threadIdx_x // 49 * 144 + 45]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 405] * kernel_shared_1[threadIdx_x // 49 * 144 + 117]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 406] * kernel_shared_1[threadIdx_x // 49 * 144 + 46]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 406] * kernel_shared_1[threadIdx_x // 49 * 144 + 118]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 407] * kernel_shared_1[threadIdx_x // 49 * 144 + 47]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 407] * kernel_shared_1[threadIdx_x // 49 * 144 + 119]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 486] * kernel_shared_1[threadIdx_x // 49 * 144 + 54]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 486] * kernel_shared_1[threadIdx_x // 49 * 144 + 126]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 487] * kernel_shared_1[threadIdx_x // 49 * 144 + 55]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 487] * kernel_shared_1[threadIdx_x // 49 * 144 + 127]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 488] * kernel_shared_1[threadIdx_x // 49 * 144 + 56]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 488] * kernel_shared_1[threadIdx_x // 49 * 144 + 128]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 567] * kernel_shared_1[threadIdx_x // 49 * 144 + 63]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 567] * kernel_shared_1[threadIdx_x // 49 * 144 + 135]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 568] * kernel_shared_1[threadIdx_x // 49 * 144 + 64]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 568] * kernel_shared_1[threadIdx_x // 49 * 144 + 136]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 569] * kernel_shared_1[threadIdx_x // 49 * 144 + 65]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 569] * kernel_shared_1[threadIdx_x // 49 * 144 + 137]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 49 * 144 + 3]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 49 * 144 + 75]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 49 * 144 + 4]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 49 * 144 + 76]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 49 * 144 + 5]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 49 * 144 + 77]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 49 * 144 + 12]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 49 * 144 + 84]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 49 * 144 + 13]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 49 * 144 + 85]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 49 * 144 + 14]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 49 * 144 + 86]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 171] * kernel_shared_1[threadIdx_x // 49 * 144 + 21]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 171] * kernel_shared_1[threadIdx_x // 49 * 144 + 93]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 172] * kernel_shared_1[threadIdx_x // 49 * 144 + 22]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 172] * kernel_shared_1[threadIdx_x // 49 * 144 + 94]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 173] * kernel_shared_1[threadIdx_x // 49 * 144 + 23]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 173] * kernel_shared_1[threadIdx_x // 49 * 144 + 95]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 252] * kernel_shared_1[threadIdx_x // 49 * 144 + 30]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 252] * kernel_shared_1[threadIdx_x // 49 * 144 + 102]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 253] * kernel_shared_1[threadIdx_x // 49 * 144 + 31]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 253] * kernel_shared_1[threadIdx_x // 49 * 144 + 103]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 254] * kernel_shared_1[threadIdx_x // 49 * 144 + 32]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 254] * kernel_shared_1[threadIdx_x // 49 * 144 + 104]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 333] * kernel_shared_1[threadIdx_x // 49 * 144 + 39]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 333] * kernel_shared_1[threadIdx_x // 49 * 144 + 111]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 334] * kernel_shared_1[threadIdx_x // 49 * 144 + 40]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 334] * kernel_shared_1[threadIdx_x // 49 * 144 + 112]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 335] * kernel_shared_1[threadIdx_x // 49 * 144 + 41]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 335] * kernel_shared_1[threadIdx_x // 49 * 144 + 113]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 414] * kernel_shared_1[threadIdx_x // 49 * 144 + 48]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 414] * kernel_shared_1[threadIdx_x // 49 * 144 + 120]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 415] * kernel_shared_1[threadIdx_x // 49 * 144 + 49]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 415] * kernel_shared_1[threadIdx_x // 49 * 144 + 121]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 416] * kernel_shared_1[threadIdx_x // 49 * 144 + 50]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 416] * kernel_shared_1[threadIdx_x // 49 * 144 + 122]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 495] * kernel_shared_1[threadIdx_x // 49 * 144 + 57]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 495] * kernel_shared_1[threadIdx_x // 49 * 144 + 129]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 496] * kernel_shared_1[threadIdx_x // 49 * 144 + 58]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 496] * kernel_shared_1[threadIdx_x // 49 * 144 + 130]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 497] * kernel_shared_1[threadIdx_x // 49 * 144 + 59]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 497] * kernel_shared_1[threadIdx_x // 49 * 144 + 131]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 576] * kernel_shared_1[threadIdx_x // 49 * 144 + 66]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 576] * kernel_shared_1[threadIdx_x // 49 * 144 + 138]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 577] * kernel_shared_1[threadIdx_x // 49 * 144 + 67]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 577] * kernel_shared_1[threadIdx_x // 49 * 144 + 139]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 578] * kernel_shared_1[threadIdx_x // 49 * 144 + 68]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 578] * kernel_shared_1[threadIdx_x // 49 * 144 + 140]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 49 * 144 + 6]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 49 * 144 + 78]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 49 * 144 + 7]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 49 * 144 + 79]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 49 * 144 + 8]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 49 * 144 + 80]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 49 * 144 + 15]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 49 * 144 + 87]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 49 * 144 + 16]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 49 * 144 + 88]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 49 * 144 + 17]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 49 * 144 + 89]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 180] * kernel_shared_1[threadIdx_x // 49 * 144 + 24]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 180] * kernel_shared_1[threadIdx_x // 49 * 144 + 96]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 181] * kernel_shared_1[threadIdx_x // 49 * 144 + 25]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 181] * kernel_shared_1[threadIdx_x // 49 * 144 + 97]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 182] * kernel_shared_1[threadIdx_x // 49 * 144 + 26]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 182] * kernel_shared_1[threadIdx_x // 49 * 144 + 98]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 261] * kernel_shared_1[threadIdx_x // 49 * 144 + 33]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 261] * kernel_shared_1[threadIdx_x // 49 * 144 + 105]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 262] * kernel_shared_1[threadIdx_x // 49 * 144 + 34]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 262] * kernel_shared_1[threadIdx_x // 49 * 144 + 106]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 263] * kernel_shared_1[threadIdx_x // 49 * 144 + 35]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 263] * kernel_shared_1[threadIdx_x // 49 * 144 + 107]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 342] * kernel_shared_1[threadIdx_x // 49 * 144 + 42]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 342] * kernel_shared_1[threadIdx_x // 49 * 144 + 114]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 343] * kernel_shared_1[threadIdx_x // 49 * 144 + 43]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 343] * kernel_shared_1[threadIdx_x // 49 * 144 + 115]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 344] * kernel_shared_1[threadIdx_x // 49 * 144 + 44]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 344] * kernel_shared_1[threadIdx_x // 49 * 144 + 116]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 423] * kernel_shared_1[threadIdx_x // 49 * 144 + 51]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 423] * kernel_shared_1[threadIdx_x // 49 * 144 + 123]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 424] * kernel_shared_1[threadIdx_x // 49 * 144 + 52]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 424] * kernel_shared_1[threadIdx_x // 49 * 144 + 124]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 425] * kernel_shared_1[threadIdx_x // 49 * 144 + 53]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 425] * kernel_shared_1[threadIdx_x // 49 * 144 + 125]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 504] * kernel_shared_1[threadIdx_x // 49 * 144 + 60]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 504] * kernel_shared_1[threadIdx_x // 49 * 144 + 132]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 505] * kernel_shared_1[threadIdx_x // 49 * 144 + 61]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 505] * kernel_shared_1[threadIdx_x // 49 * 144 + 133]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 506] * kernel_shared_1[threadIdx_x // 49 * 144 + 62]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 506] * kernel_shared_1[threadIdx_x // 49 * 144 + 134]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 585] * kernel_shared_1[threadIdx_x // 49 * 144 + 69]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 585] * kernel_shared_1[threadIdx_x // 49 * 144 + 141]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 586] * kernel_shared_1[threadIdx_x // 49 * 144 + 70]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 586] * kernel_shared_1[threadIdx_x // 49 * 144 + 142]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 587] * kernel_shared_1[threadIdx_x // 49 * 144 + 71]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 49 // 7 * 9 + threadIdx_x % 7 + 587] * kernel_shared_1[threadIdx_x // 49 * 144 + 143]
- for i1_inner in range(2):
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 28] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 28) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 56] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 56) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 84] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 84) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 112] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 112) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 140] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 140) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 168] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 168) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 196] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 196) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 224) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 252] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 32256]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 280] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 280) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 308] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 308) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 336] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 336) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 364] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 364) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 392] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 392) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 420] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 420) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 448) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 476] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 476) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 504] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 64512]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 532] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 532) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 560] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 560) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 588] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 588) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 616] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 616) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 644] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 644) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 672) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 700] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 700) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 728] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 728) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 756] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 96768]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 784) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 812] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 812) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 840] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 840) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 868] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 868) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 896) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 924] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 924) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 952] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 952) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 980] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 980) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1008] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 129024]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1036] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1036) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1064] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1064) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1092] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1092) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1120) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1148] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1148) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1176] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1176) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1204] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1204) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1232] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1232) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1260] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 161280]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1288] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1288) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1316] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1316) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1344) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1372] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1372) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1400] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1400) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1428] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1428) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1456] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1456) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1484] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1484) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1512] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 193536]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1540] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1540) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1568) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1596] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1596) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1624] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1624) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1652] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1652) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1680] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1680) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1708] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1708) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1736] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1736) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1764] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 225792]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1792) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1820] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1820) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1848] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1848) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1876] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1876) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1904] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1904) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1932] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1932) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1960] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1960) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 1988] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 1988) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 258048]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2044] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2044) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 28) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2072] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2072) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 20) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2100] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2100) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 12) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2128] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2128) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 4]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2156] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2156) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2184] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2184) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 24) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2212] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2212) // 36 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 36]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2240] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2240) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 8]
+ with T.launch_thread(threadIdx_x_2, 28):
+ kernel_shared_1[threadIdx_x_2 + 2268] = kernel_1[blockIdx_x // 7 * 294912 + cse_var_1 + threadIdx_x_2 + 290304]
+ with T.launch_thread(threadIdx_x_2, 28):
+ if T.likely(threadIdx_x_2 < 8):
+ kernel_shared_1[threadIdx_x_2 + 2296] = kernel_1[blockIdx_x // 7 * 294912 + (threadIdx_x_2 + 2296) // 36 * 4608 + cse_var_1 + threadIdx_x_2 + 28]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 576]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1152]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1728]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 36]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 612]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1188]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1764]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 577]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1153]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1729]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 37]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 613]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1189]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1765]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 2]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 578]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1154]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1730]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 38]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 614]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1190]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1766]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 3]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 579]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1155]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1731]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 39]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 615]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1191]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1767]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 4]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 580]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1156]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1732]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 40]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 616]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1192]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1768]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 5]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 581]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1157]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1733]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 41]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 617]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1193]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1769]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 6]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 582]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1158]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1734]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 42]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 618]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1194]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1770]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 7]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 583]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1159]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1735]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 43]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 619]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1195]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1771]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 8]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 584]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1160]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1736]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 44]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 620]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1196]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1772]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 72]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 648]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1224]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1800]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 108]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 684]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1260]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7] * kernel_shared_1[threadIdx_x // 7 * 144 + 1836]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 73]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 649]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1225]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1801]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 109]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 685]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1261]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 144 + 1837]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 74]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 650]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1226]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1802]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 110]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 686]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1262]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 144 + 1838]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 75]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 651]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1227]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1803]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 111]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 687]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1263]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 9] * kernel_shared_1[threadIdx_x // 7 * 144 + 1839]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 76]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 652]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1228]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1804]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 112]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 688]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1264]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 10] * kernel_shared_1[threadIdx_x // 7 * 144 + 1840]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 77]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 653]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1229]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1805]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 113]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 689]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1265]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 11] * kernel_shared_1[threadIdx_x // 7 * 144 + 1841]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 78]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 654]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1230]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1806]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 114]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 690]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1266]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 18] * kernel_shared_1[threadIdx_x // 7 * 144 + 1842]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 79]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 655]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1231]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1807]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 115]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 691]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1267]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 19] * kernel_shared_1[threadIdx_x // 7 * 144 + 1843]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 80]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 656]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1232]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1808]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 116]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 692]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1268]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 20] * kernel_shared_1[threadIdx_x // 7 * 144 + 1844]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 9]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 585]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1161]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1737]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 45]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 621]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1197]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1773]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 10]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 586]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1162]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1738]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 46]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 622]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1198]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1774]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 11]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 587]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1163]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1739]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 47]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 623]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1199]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1775]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 12]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 588]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1164]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1740]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 48]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 624]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1200]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1776]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 13]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 589]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1165]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1741]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 49]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 625]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1201]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1777]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 14]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 590]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1166]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1742]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 50]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 626]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1202]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1778]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 15]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 591]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1167]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1743]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 51]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 627]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1203]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1779]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 16]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 592]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1168]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1744]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 52]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 628]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1204]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1780]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 17]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 593]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1169]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1745]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 53]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 629]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1205]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1781]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 81]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 657]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1233]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1809]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 117]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 693]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1269]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 27] * kernel_shared_1[threadIdx_x // 7 * 144 + 1845]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 82]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 658]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1234]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1810]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 118]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 694]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1270]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 28] * kernel_shared_1[threadIdx_x // 7 * 144 + 1846]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 83]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 659]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1235]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1811]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 119]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 695]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1271]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 29] * kernel_shared_1[threadIdx_x // 7 * 144 + 1847]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 84]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 660]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1236]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1812]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 120]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 696]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1272]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 36] * kernel_shared_1[threadIdx_x // 7 * 144 + 1848]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 85]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 661]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1237]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1813]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 121]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 697]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1273]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 37] * kernel_shared_1[threadIdx_x // 7 * 144 + 1849]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 86]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 662]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1238]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1814]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 122]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 698]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1274]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 38] * kernel_shared_1[threadIdx_x // 7 * 144 + 1850]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 87]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 663]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1239]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1815]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 123]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 699]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1275]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 45] * kernel_shared_1[threadIdx_x // 7 * 144 + 1851]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 88]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 664]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1240]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1816]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 124]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 700]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1276]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 46] * kernel_shared_1[threadIdx_x // 7 * 144 + 1852]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 89]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 665]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1241]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1817]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 125]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 701]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1277]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 47] * kernel_shared_1[threadIdx_x // 7 * 144 + 1853]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 18]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 594]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1170]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1746]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 54]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 630]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1206]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1782]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 19]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 595]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1171]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1747]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 55]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 631]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1207]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1783]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 20]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 596]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1172]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1748]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 56]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 632]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1208]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1784]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 21]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 597]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1173]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1749]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 57]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 633]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1209]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1785]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 22]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 598]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1174]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1750]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 58]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 634]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1210]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1786]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 23]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 599]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1175]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1751]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 59]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 635]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1211]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1787]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 24]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 600]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1176]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1752]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 60]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 636]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1212]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1788]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 25]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 601]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1177]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1753]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 61]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 637]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1213]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1789]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 26]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 602]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1178]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1754]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 62]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 638]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1214]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1790]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 90]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 666]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1242]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1818]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 126]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 702]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1278]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 144 + 1854]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 91]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 667]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1243]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1819]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 127]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 703]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1279]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 144 + 1855]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 92]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 668]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1244]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1820]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 128]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 704]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1280]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 56] * kernel_shared_1[threadIdx_x // 7 * 144 + 1856]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 93]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 669]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1245]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1821]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 129]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 705]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1281]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 144 + 1857]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 94]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 670]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1246]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1822]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 130]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 706]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1282]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 144 + 1858]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 95]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 671]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1247]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1823]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 131]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 707]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1283]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 144 + 1859]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 96]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 672]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1248]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1824]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 132]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 708]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1284]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 72] * kernel_shared_1[threadIdx_x // 7 * 144 + 1860]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 97]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 673]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1249]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1825]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 133]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 709]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1285]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 73] * kernel_shared_1[threadIdx_x // 7 * 144 + 1861]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 98]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 674]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1250]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1826]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 134]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 710]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1286]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 74] * kernel_shared_1[threadIdx_x // 7 * 144 + 1862]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 27]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 603]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1179]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1755]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 63]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 639]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1215]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1791]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 28]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 604]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1180]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1756]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 64]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 640]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1216]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1792]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 29]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 605]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1181]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1757]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 65]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 641]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1217]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1793]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 30]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 606]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1182]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1758]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 66]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 642]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1218]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1794]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 31]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 607]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1183]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1759]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 67]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 643]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1219]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1795]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 32]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 608]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1184]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1760]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 68]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 644]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1220]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1796]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 33]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 609]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1185]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1761]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 69]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 645]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1221]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1797]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 34]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 610]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1186]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1762]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 70]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 646]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1222]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1798]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 35]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 611]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1187]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1763]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 71]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 647]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1223]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1799]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 99]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 675]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1251]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1827]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 135]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 711]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1287]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 81] * kernel_shared_1[threadIdx_x // 7 * 144 + 1863]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 100]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 676]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1252]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1828]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 136]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 712]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1288]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 82] * kernel_shared_1[threadIdx_x // 7 * 144 + 1864]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 101]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 677]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1253]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1829]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 137]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 713]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1289]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 83] * kernel_shared_1[threadIdx_x // 7 * 144 + 1865]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 102]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 678]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1254]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1830]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 138]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 714]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1290]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 90] * kernel_shared_1[threadIdx_x // 7 * 144 + 1866]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 103]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 679]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1255]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1831]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 139]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 715]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1291]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 91] * kernel_shared_1[threadIdx_x // 7 * 144 + 1867]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 104]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 680]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1256]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1832]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 140]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 716]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1292]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 92] * kernel_shared_1[threadIdx_x // 7 * 144 + 1868]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 105]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 681]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1257]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1833]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 141]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 717]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1293]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 144 + 1869]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 106]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 682]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1258]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1834]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 142]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 718]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1294]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 144 + 1870]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 107]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 683]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1259]
+ conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1835]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 143]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 719]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1295]
+ conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[threadIdx_x % 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 144 + 1871]
+ for i1_inner in range(4):
compute_1 = T.Buffer((25088,), data=compute.data)
bias_1 = T.Buffer((512,), data=bias.data)
- compute_1[blockIdx_x * 1568 + threadIdx_x // 49 * 98 + i1_inner * 49 + threadIdx_x % 49] = T.max(conv2d_nchw_1[i1_inner] + bias_1[blockIdx_x * 32 + threadIdx_x // 49 * 2 + i1_inner], T.float32(0))
+ compute_1[blockIdx_x // 7 * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + blockIdx_x % 7 * 7 + threadIdx_x % 7] = T.max(conv2d_nchw_1[i1_inner] + bias_1[blockIdx_x // 7 * 64 + threadIdx_x // 7 * 4 + i1_inner], T.float32(0))
+ compute_1[blockIdx_x // 7 * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + blockIdx_x % 7 * 7 + threadIdx_x % 7 + 784] = T.max(conv2d_nchw_1[i1_inner + 4] + bias_1[blockIdx_x // 7 * 64 + threadIdx_x // 7 * 4 + i1_inner + 16], T.float32(0))
+ compute_1[blockIdx_x // 7 * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + blockIdx_x % 7 * 7 + threadIdx_x % 7 + 1568] = T.max(conv2d_nchw_1[i1_inner + 8] + bias_1[blockIdx_x // 7 * 64 + threadIdx_x // 7 * 4 + i1_inner + 32], T.float32(0))
+ compute_1[blockIdx_x // 7 * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + blockIdx_x % 7 * 7 + threadIdx_x % 7 + 2352] = T.max(conv2d_nchw_1[i1_inner + 12] + bias_1[blockIdx_x // 7 * 64 + threadIdx_x // 7 * 4 + i1_inner + 48], T.float32(0))
</pre></div>
</div>
</div>
@@ -714,7 +1330,7 @@ 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: 0.383 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.357 ms
</pre></div>
</div>
</div>
@@ -744,32 +1360,32 @@ conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o
conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=2)
-conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=1)
-conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=16)
-conv2d_nchw_ff_o_o_o_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_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=2)
+conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=4)
+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=4)
conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
-conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=7)
+conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=1)
conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=1)
conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=1)
conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=7)
conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=1)
-conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=8)
-conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=1)
-conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
-conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=3)
+conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=1)
+conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=4)
+conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=3)
+conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=3)
conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=1)
s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2d_nc [...]
compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
-compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=2)
-compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=16)
-compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=4)
+compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=4)
+compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=4)
compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
-compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
+compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=1)
compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=7)
@@ -792,14 +1408,14 @@ s[compute].bind(compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused, te.thread
kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[kernel_shared].fuse(kernel_shared_ax0, kernel_shared_ax1, kernel_shared_ax2, kernel_shared_ax3)
kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
s[kernel_shared].vectorize(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
-kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=784)
+kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=28)
s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[pad_temp_shared].fuse(pad_temp_shared_ax0, pad_temp_shared_ax1, pad_temp_shared_ax2, pad_temp_shared_ax3)
pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
-pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=784)
+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=28)
s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
-s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
+s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 1024)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
CUDA source code:
@@ -817,170 +1433,702 @@ CUDA source code:
#define int64_t long long
#define uint64_t unsigned long long
#endif
-extern "C" __global__ void __launch_bounds__(784) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[2];
- __shared__ float pad_temp_shared[648];
+extern "C" __global__ void __launch_bounds__(28) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
+ float conv2d_nchw[16];
+ __shared__ float pad_temp_shared[108];
__shared__ float kernel_shared[2304];
conv2d_nchw[0] = 0.000000e+00f;
+ conv2d_nchw[4] = 0.000000e+00f;
+ conv2d_nchw[8] = 0.000000e+00f;
+ conv2d_nchw[12] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
- for (int rc_outer_outer = 0; rc_outer_outer < 64; ++rc_outer_outer) {
+ conv2d_nchw[5] = 0.000000e+00f;
+ conv2d_nchw[9] = 0.000000e+00f;
+ conv2d_nchw[13] = 0.000000e+00f;
+ conv2d_nchw[2] = 0.000000e+00f;
+ conv2d_nchw[6] = 0.000000e+00f;
+ conv2d_nchw[10] = 0.000000e+00f;
+ conv2d_nchw[14] = 0.000000e+00f;
+ conv2d_nchw[3] = 0.000000e+00f;
+ conv2d_nchw[7] = 0.000000e+00f;
+ conv2d_nchw[11] = 0.000000e+00f;
+ conv2d_nchw[15] = 0.000000e+00f;
+ for (int rc_outer_outer = 0; rc_outer_outer < 128; ++rc_outer_outer) {
__syncthreads();
- if (((int)threadIdx.x) < 648) {
- pad_temp_shared[((int)threadIdx.x)] = (((((9 <= (((int)threadIdx.x) % 81)) && ((((int)threadIdx.x) % 81) < 72)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 81) * 49)) + (((((int)threadIdx.x) % 81) / 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[((int)threadIdx.x)] = (((((1 <= (((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7))) && ((((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((((rc_outer_outer * 196) + ((((int)threadIdx.x) / 27) * 49)) + (((((int)threadIdx.x) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 28)] = (((((1 <= ((((((int)threadIdx.x) + 1) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 1) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 28) / 27) * 49)) + ((((((int)threadIdx.x) + 1) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)thr [...]
+ pad_temp_shared[(((int)threadIdx.x) + 56)] = (((((1 <= ((((((int)threadIdx.x) + 2) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 2) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 56) / 27) * 49)) + ((((((int)threadIdx.x) + 2) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)thr [...]
+ if (((int)threadIdx.x) < 24) {
+ pad_temp_shared[(((int)threadIdx.x) + 84)] = (((((1 <= (((((int)threadIdx.x) + 3) / 9) + (((int)blockIdx.x) % 7))) && ((((((int)threadIdx.x) + 3) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[((((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 84) / 27) * 49)) + (((((int)threadIdx.x) + 3) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 3) % 9) [...]
}
- kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) * 147456) + ((((int)threadIdx.x) / 72) * 4608)) + (rc_outer_outer * 72)) + (((int)threadIdx.x) % 72))];
- kernel_shared[(((int)threadIdx.x) + 784)] = kernel[(((((((int)blockIdx.x) * 147456) + (((((int)threadIdx.x) + 784) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- if (((int)threadIdx.x) < 736) {
- kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[(((((((int)blockIdx.x) * 147456) + (((((int)threadIdx.x) + 1568) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x))];
+ kernel_shared[(((int)threadIdx.x) + 28)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 28) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 56)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 56) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 84)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 84) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 112)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 112) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 140)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 140) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 168)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 168) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 196)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 196) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 224)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 224) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 252)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 32256)];
+ kernel_shared[(((int)threadIdx.x) + 280)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 280) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 308)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 308) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 336)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 336) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 364)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 364) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 392)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 392) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 420)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 420) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 448) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 476)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 476) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 504)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 64512)];
+ kernel_shared[(((int)threadIdx.x) + 532)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 532) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 560)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 560) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 588)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 588) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 616)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 616) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 644)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 644) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 672)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 672) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 700)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 700) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 728)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 728) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 756)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 96768)];
+ kernel_shared[(((int)threadIdx.x) + 784)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 784) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 812)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 812) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 840)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 840) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 868)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 868) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 896) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 924)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 924) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 952)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 952) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 980)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 980) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 1008)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 129024)];
+ kernel_shared[(((int)threadIdx.x) + 1036)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1036) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1064)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1064) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1092)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1092) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1120) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 1148)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1148) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1176)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1176) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1204)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1204) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1232)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1232) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 1260)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 161280)];
+ kernel_shared[(((int)threadIdx.x) + 1288)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1288) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1316)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1316) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1344) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1372)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1372) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 1400)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1400) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1428)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1428) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1456)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1456) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1484)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1484) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 1512)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 193536)];
+ kernel_shared[(((int)threadIdx.x) + 1540)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1540) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1568) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1596)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1596) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1624)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1624) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 1652)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1652) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1680)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1680) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1708)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1708) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1736)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1736) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 1764)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 225792)];
+ kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1792) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1820)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1820) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1848)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1848) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1876)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1876) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 1904)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1904) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1932)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1932) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1960)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1960) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 1988)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1988) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 258048)];
+ kernel_shared[(((int)threadIdx.x) + 2044)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2044) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 28) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2072)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2072) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 20) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2100)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2100) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 12) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2128)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2128) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 4)];
+ kernel_shared[(((int)threadIdx.x) + 2156)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2156) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 32) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2184)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2184) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 24) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2212)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2212) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((int)threadIdx.x) + 16) % 36))];
+ kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2240) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 8)];
+ kernel_shared[(((int)threadIdx.x) + 2268)] = kernel[(((((((int)blockIdx.x) / 7) * 294912) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 290304)];
+ if (((int)threadIdx.x) < 8) {
+ kernel_shared[(((int)threadIdx.x) + 2296)] = kernel[((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 2296) / 36) * 4608)) + (rc_outer_outer * 36)) + ((int)threadIdx.x)) + 28)];
}
__syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[((((int)threadIdx.x) / 49) * 144)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 72)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 73)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 74)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 81)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 82)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 83)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 90)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 91)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 92)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 27)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 99)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 28)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 100)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 29)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 101)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 324)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 36)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 324)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 108)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 325)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 37)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 325)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 109)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 326)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 38)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 326)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 110)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 405)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 45)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 405)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 117)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 406)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 46)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 406)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 118)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 407)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 47)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 407)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 119)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 486)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 54)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 486)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 126)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 487)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 55)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 487)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 127)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 488)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 56)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 488)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 128)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 567)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 63)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 567)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 135)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 568)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 64)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 568)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 136)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 569)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 65)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 569)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 137)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 75)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 76)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 77)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 84)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 85)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 86)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 93)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 94)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 95)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 30)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 102)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 31)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 103)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 32)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 104)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 333)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 39)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 333)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 111)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 334)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 40)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 334)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 112)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 335)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 41)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 335)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 113)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 414)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 48)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 414)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 120)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 415)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 49)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 415)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 121)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 416)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 50)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 416)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 122)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 495)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 57)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 495)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 129)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 496)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 58)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 496)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 130)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 497)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 59)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 497)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 131)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 576)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 66)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 576)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 138)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 577)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 67)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 577)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 139)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 578)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 68)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 578)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 140)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 78)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 79)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 80)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 87)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 88)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 89)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 96)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 25)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 97)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 26)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 98)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 33)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 105)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 34)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 106)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 35)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 107)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 342)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 42)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 342)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 114)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 343)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 43)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 343)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 115)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 344)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 44)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 344)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 116)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 423)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 51)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 423)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 123)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 424)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 52)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 424)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 124)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 425)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 53)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 425)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 125)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 504)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 60)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 504)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 132)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 505)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 61)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 505)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 133)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 506)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 62)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 506)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 134)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 585)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 69)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 585)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 141)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 586)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 70)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 586)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 142)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 587)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 71)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((((int)threadIdx.x) % 49) / 7) * 9) + (((int)threadIdx.x) % 7)) + 587)] * kernel_shared[(((((int)threadIdx.x) / 49) * 144) + 143)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[((((int)threadIdx.x) / 7) * 144)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 576)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1152)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1728)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 36)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 612)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1188)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1764)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 577)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1153)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1729)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 37)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 613)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1189)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1765)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 2)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 578)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1154)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1730)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 38)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 614)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1190)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1766)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 3)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 579)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1155)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1731)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 39)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 615)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1191)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1767)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 4)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 580)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1156)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1732)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 40)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 616)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1192)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1768)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 5)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 581)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1157)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1733)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 41)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 617)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1193)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1769)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 6)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 582)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1158)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1734)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 42)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 618)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1194)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1770)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 7)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 583)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1159)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1735)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 43)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 619)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1195)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1771)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 8)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 584)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1160)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1736)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 44)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 620)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1196)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1772)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 72)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 648)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1224)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1800)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 108)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 684)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1260)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) % 7)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1836)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 73)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 649)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1225)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1801)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 109)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 685)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1261)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1837)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 74)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 650)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1226)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1802)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 110)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 686)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1262)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1838)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 75)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 651)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1227)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1803)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 111)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 687)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1263)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 9)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1839)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 76)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 652)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1228)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1804)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 112)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 688)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1264)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 10)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1840)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 77)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 653)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1229)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1805)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 113)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 689)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1265)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 11)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1841)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 78)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 654)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1230)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1806)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 114)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 690)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1266)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 18)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1842)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 79)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 655)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1231)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1807)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 115)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 691)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1267)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 19)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1843)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 80)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 656)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1232)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1808)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 116)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 692)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1268)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 20)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1844)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 9)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 585)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1161)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1737)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 45)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 621)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1197)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1773)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 10)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 586)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1162)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1738)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 46)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 622)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1198)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1774)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 11)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 587)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1163)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1739)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 47)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 623)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1199)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1775)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 12)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 588)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1164)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1740)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 48)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 624)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1200)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1776)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 13)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 589)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1165)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1741)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 49)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 625)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1201)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1777)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 14)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 590)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1166)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1742)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 50)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 626)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1202)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1778)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 15)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 591)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1167)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1743)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 51)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 627)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1203)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1779)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 16)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 592)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1168)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1744)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 52)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 628)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1204)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1780)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 17)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 593)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1169)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1745)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 53)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 629)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1205)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1781)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 81)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 657)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1233)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1809)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 117)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 693)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1269)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 27)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1845)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 82)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 658)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1234)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1810)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 118)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 694)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1270)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 28)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1846)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 83)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 659)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1235)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1811)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 119)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 695)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1271)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 29)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1847)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 84)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 660)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1236)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1812)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 120)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 696)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1272)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 36)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1848)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 85)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 661)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1237)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1813)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 121)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 697)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1273)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 37)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1849)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 86)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 662)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1238)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1814)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 122)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 698)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1274)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 38)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1850)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 87)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 663)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1239)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1815)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 123)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 699)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1275)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 45)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1851)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 88)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 664)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1240)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1816)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 124)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 700)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1276)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 46)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1852)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 89)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 665)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1241)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1817)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 125)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 701)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1277)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 47)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1853)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 18)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 594)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1170)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1746)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 54)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 630)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1206)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1782)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 19)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 595)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1171)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1747)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 55)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 631)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1207)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1783)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 20)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 596)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1172)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1748)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 56)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 632)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1208)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1784)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 21)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 597)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1173)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1749)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 57)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 633)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1209)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1785)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 22)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 598)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1174)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1750)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 58)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 634)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1210)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1786)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 23)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 599)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1175)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1751)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 59)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 635)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1211)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1787)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 24)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 600)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1176)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1752)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 60)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 636)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1212)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1788)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 25)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 601)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1177)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1753)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 61)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 637)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1213)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1789)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 26)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 602)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1178)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1754)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 62)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 638)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1214)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1790)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 90)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 666)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1242)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1818)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 126)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 702)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1278)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 54)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1854)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 91)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 667)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1243)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1819)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 127)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 703)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1279)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 55)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1855)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 92)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 668)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1244)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1820)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 128)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 704)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1280)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 56)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1856)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 93)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 669)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1245)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1821)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 129)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 705)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1281)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 63)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1857)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 94)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 670)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1246)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1822)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 130)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 706)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1282)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 64)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1858)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 95)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 671)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1247)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1823)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 131)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 707)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1283)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 65)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1859)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 96)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 672)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1248)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1824)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 132)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 708)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1284)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 72)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1860)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 97)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 673)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1249)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1825)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 133)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 709)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1285)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 73)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1861)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 98)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 674)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1250)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1826)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 134)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 710)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1286)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 74)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1862)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 27)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 603)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1179)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1755)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 63)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 639)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1215)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1791)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 28)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 604)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1180)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1756)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 64)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 640)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1216)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1792)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 29)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 605)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1181)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1757)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 65)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 641)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1217)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1793)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 30)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 606)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1182)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1758)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 66)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 642)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1218)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1794)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 31)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 607)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1183)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1759)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 67)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 643)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1219)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1795)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 32)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 608)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1184)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1760)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 68)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 644)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1220)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1796)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 33)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 609)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1185)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1761)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 69)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 645)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1221)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1797)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 34)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 610)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1186)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1762)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 70)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 646)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1222)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1798)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 35)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 611)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1187)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1763)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 71)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 647)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1223)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1799)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 99)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 675)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1251)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1827)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 135)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 711)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1287)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 81)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1863)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 100)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 676)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1252)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1828)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 136)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 712)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1288)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 82)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1864)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 101)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 677)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1253)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1829)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 137)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 713)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1289)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 83)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1865)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 102)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 678)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1254)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1830)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 138)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 714)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1290)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 90)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1866)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 103)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 679)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1255)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1831)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 139)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 715)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1291)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 91)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1867)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 104)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 680)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1256)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1832)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 140)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 716)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1292)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 92)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1868)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 105)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 681)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1257)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1833)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 141)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 717)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1293)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 99)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1869)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 106)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 682)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1258)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1834)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 142)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 718)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1294)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 100)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1870)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 107)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 683)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1259)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1835)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 143)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 719)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1295)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((int)threadIdx.x) % 7) + 101)] * kernel_shared[(((((int)threadIdx.x) / 7) * 144) + 1871)]));
}
- for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
- compute[((((((int)blockIdx.x) * 1568) + ((((int)threadIdx.x) / 49) * 98)) + (i1_inner * 49)) + (((int)threadIdx.x) % 49))] = max((conv2d_nchw[i1_inner] + bias[(((((int)blockIdx.x) * 32) + ((((int)threadIdx.x) / 49) * 2)) + i1_inner)]), 0.000000e+00f);
+ for (int i1_inner = 0; i1_inner < 4; ++i1_inner) {
+ compute[((((((((int)blockIdx.x) / 7) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 7))] = max((conv2d_nchw[i1_inner] + bias[((((((int)blockIdx.x) / 7) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner)]), 0.000000e+00f);
+ compute[(((((((((int)blockIdx.x) / 7) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 7)) + 784)] = max((conv2d_nchw[(i1_inner + 4)] + bias[(((((((int)blockIdx.x) / 7) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner) + 16)]), 0.000000e+00f);
+ compute[(((((((((int)blockIdx.x) / 7) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 7)) + 1568)] = max((conv2d_nchw[(i1_inner + 8)] + bias[(((((((int)blockIdx.x) / 7) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner) + 32)]), 0.000000e+00f);
+ compute[(((((((((int)blockIdx.x) / 7) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 7)) + 2352)] = max((conv2d_nchw[(i1_inner + 12)] + bias[(((((((int)blockIdx.x) / 7) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner) + 48)]), 0.000000e+00f);
}
}
</pre></div>
@@ -1017,7 +2165,7 @@ In the example below we resume the status and do more 5 trials.</p>
Get devices for measurement successfully!
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 5 minutes 59.876 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 6 minutes 8.809 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 84e80ae8d9..9e361ecb2e 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
@@ -916,7 +916,7 @@ so we can read the log file and load the best schedules.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 7.9015 7.8961 7.9151 7.8933 0.0097
+ 7.9351 7.9334 7.9456 7.9262 0.0080
</pre></div>
</div>
</div>
@@ -938,7 +938,7 @@ to learn how to use the RPC Tracker and RPC Server.
To use the RPC Tracker in auto-scheduler, replace the runner in <code class="code docutils literal notranslate"><span class="pre">TuningOptions</span></code>
with <a class="reference internal" href="../../reference/api/python/auto_scheduler.html#tvm.auto_scheduler.RPCRunner" title="tvm.auto_scheduler.RPCRunner"><code class="xref any py py-class docutils literal notranslate"><span class="pre">auto_scheduler.RPCRunner</span></code></a>.</p></li>
</ol>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 7.619 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 7.165 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 f11acc2021..710c48710b 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
@@ -935,7 +935,7 @@ so we can read the log file and load the best schedules.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 748.8999 748.8494 749.0310 748.8193 0.0935
+ 760.6615 760.3228 762.8023 758.8594 1.6274
</pre></div>
</div>
</div>
@@ -957,7 +957,7 @@ to learn how to use the RPC Tracker and RPC Server.
To use the RPC Tracker in auto-scheduler, replace the runner in <code class="code docutils literal notranslate"><span class="pre">TuningOptions</span></code>
with <a class="reference internal" href="../../reference/api/python/auto_scheduler.html#tvm.auto_scheduler.RPCRunner" title="tvm.auto_scheduler.RPCRunner"><code class="xref any py py-class docutils literal notranslate"><span class="pre">auto_scheduler.RPCRunner</span></code></a>.</p></li>
</ol>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 40.136 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 40.129 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 8980823ebc..8d63997677 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
@@ -632,87 +632,404 @@ class Module:
@T.prim_func
def main(placeholder: T.Buffer((128, 256), "float32"), placeholder_1: T.Buffer((4916, 16, 1), "float32"), placeholder_2: T.Buffer((4916,), "int32"), placeholder_3: T.Buffer((33,), "int32"), placeholder_4: T.Buffer((128, 512), "float32"), compute: T.Buffer((128, 512), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
- for i0_outer in T.parallel(4):
- compute_1 = T.allocate([512], "float32", "global")
- for i1_outer in range(32):
- compute_2 = T.Buffer((512,), data=compute_1)
- for i_outer_inner in range(4):
- for i_inner_init in range(8):
- cse_var_1: T.int32 = i_outer_inner * 128 + i_inner_init * 16
- compute_2[cse_var_1] = T.float32(0)
- compute_2[cse_var_1 + 1] = T.float32(0)
- compute_2[cse_var_1 + 2] = T.float32(0)
- compute_2[cse_var_1 + 3] = T.float32(0)
- compute_2[cse_var_1 + 4] = T.float32(0)
- compute_2[cse_var_1 + 5] = T.float32(0)
- compute_2[cse_var_1 + 6] = T.float32(0)
- compute_2[cse_var_1 + 7] = T.float32(0)
- compute_2[cse_var_1 + 8] = T.float32(0)
- compute_2[cse_var_1 + 9] = T.float32(0)
- compute_2[cse_var_1 + 10] = T.float32(0)
- compute_2[cse_var_1 + 11] = T.float32(0)
- compute_2[cse_var_1 + 12] = T.float32(0)
- compute_2[cse_var_1 + 13] = T.float32(0)
- compute_2[cse_var_1 + 14] = T.float32(0)
- compute_2[cse_var_1 + 15] = T.float32(0)
- for elem_idx, i_inner in T.grid(placeholder_5[i1_outer + 1] - placeholder_5[i1_outer], 8):
- placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
- 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)
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_2: T.int32 = i_outer_inner * 128 + i_inner * 16
- compute_2[cse_var_2] = compute_2[cse_var_2] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_3: T.int32 = i_outer_inner * 128 + i_inner * 16 + 1
- compute_2[cse_var_3] = compute_2[cse_var_3] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_4: T.int32 = i_outer_inner * 128 + i_inner * 16 + 2
- compute_2[cse_var_4] = compute_2[cse_var_4] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_5: T.int32 = i_outer_inner * 128 + i_inner * 16 + 3
- compute_2[cse_var_5] = compute_2[cse_var_5] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_6: T.int32 = i_outer_inner * 128 + i_inner * 16 + 4
- compute_2[cse_var_6] = compute_2[cse_var_6] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_7: T.int32 = i_outer_inner * 128 + i_inner * 16 + 5
- compute_2[cse_var_7] = compute_2[cse_var_7] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_8: T.int32 = i_outer_inner * 128 + i_inner * 16 + 6
- compute_2[cse_var_8] = compute_2[cse_var_8] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_9: T.int32 = i_outer_inner * 128 + i_inner * 16 + 7
- compute_2[cse_var_9] = compute_2[cse_var_9] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_10: T.int32 = i_outer_inner * 128 + i_inner * 16 + 8
- compute_2[cse_var_10] = compute_2[cse_var_10] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_11: T.int32 = i_outer_inner * 128 + i_inner * 16 + 9
- compute_2[cse_var_11] = compute_2[cse_var_11] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_12: T.int32 = i_outer_inner * 128 + i_inner * 16 + 10
- compute_2[cse_var_12] = compute_2[cse_var_12] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_13: T.int32 = i_outer_inner * 128 + i_inner * 16 + 11
- compute_2[cse_var_13] = compute_2[cse_var_13] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_14: T.int32 = i_outer_inner * 128 + i_inner * 16 + 12
- compute_2[cse_var_14] = compute_2[cse_var_14] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_15: T.int32 = i_outer_inner * 128 + i_inner * 16 + 13
- compute_2[cse_var_15] = compute_2[cse_var_15] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_16: T.int32 = i_outer_inner * 128 + i_inner * 16 + 14
- compute_2[cse_var_16] = compute_2[cse_var_16] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- cse_var_17: T.int32 = i_outer_inner * 128 + i_inner * 16 + 15
- compute_2[cse_var_17] = compute_2[cse_var_17] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer * 8192 + i_outer_inner * 2048 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- for i0_inner in range(32):
- cse_var_18: T.int32 = i0_outer * 16384 + i0_inner * 512 + i1_outer * 16
- compute_3 = T.Buffer((65536,), data=compute.data)
- placeholder_5 = T.Buffer((65536,), data=placeholder_4.data)
- compute_3[cse_var_18:cse_var_18 + 16] = T.max(compute_2[i0_inner * 16:i0_inner * 16 + 16] + placeholder_5[cse_var_18:cse_var_18 + 16], T.Broadcast(T.float32(0), 16))
+ for i0_outer_i1_outer_fused in T.parallel(512):
+ cse_var_1: T.int32 = i0_outer_i1_outer_fused % 32
+ compute_1 = T.allocate([128], "float32", "global")
+ compute_2 = T.Buffer((128,), data=compute_1)
+ compute_2[0] = T.float32(0)
+ compute_2[1] = T.float32(0)
+ compute_2[2] = T.float32(0)
+ compute_2[3] = T.float32(0)
+ compute_2[4] = T.float32(0)
+ compute_2[5] = T.float32(0)
+ compute_2[6] = T.float32(0)
+ compute_2[7] = T.float32(0)
+ compute_2[8] = T.float32(0)
+ compute_2[9] = T.float32(0)
+ compute_2[10] = T.float32(0)
+ compute_2[11] = T.float32(0)
+ compute_2[12] = T.float32(0)
+ compute_2[13] = T.float32(0)
+ compute_2[14] = T.float32(0)
+ compute_2[15] = T.float32(0)
+ compute_2[16] = T.float32(0)
+ compute_2[17] = T.float32(0)
+ compute_2[18] = T.float32(0)
+ compute_2[19] = T.float32(0)
+ compute_2[20] = T.float32(0)
+ compute_2[21] = T.float32(0)
+ compute_2[22] = T.float32(0)
+ compute_2[23] = T.float32(0)
+ compute_2[24] = T.float32(0)
+ compute_2[25] = T.float32(0)
+ compute_2[26] = T.float32(0)
+ compute_2[27] = T.float32(0)
+ compute_2[28] = T.float32(0)
+ compute_2[29] = T.float32(0)
+ compute_2[30] = T.float32(0)
+ compute_2[31] = T.float32(0)
+ compute_2[32] = T.float32(0)
+ compute_2[33] = T.float32(0)
+ compute_2[34] = T.float32(0)
+ compute_2[35] = T.float32(0)
+ compute_2[36] = T.float32(0)
+ compute_2[37] = T.float32(0)
+ compute_2[38] = T.float32(0)
+ compute_2[39] = T.float32(0)
+ compute_2[40] = T.float32(0)
+ compute_2[41] = T.float32(0)
+ compute_2[42] = T.float32(0)
+ compute_2[43] = T.float32(0)
+ compute_2[44] = T.float32(0)
+ compute_2[45] = T.float32(0)
+ compute_2[46] = T.float32(0)
+ compute_2[47] = T.float32(0)
+ compute_2[48] = T.float32(0)
+ compute_2[49] = T.float32(0)
+ compute_2[50] = T.float32(0)
+ compute_2[51] = T.float32(0)
+ compute_2[52] = T.float32(0)
+ compute_2[53] = T.float32(0)
+ compute_2[54] = T.float32(0)
+ compute_2[55] = T.float32(0)
+ compute_2[56] = T.float32(0)
+ compute_2[57] = T.float32(0)
+ compute_2[58] = T.float32(0)
+ compute_2[59] = T.float32(0)
+ compute_2[60] = T.float32(0)
+ compute_2[61] = T.float32(0)
+ compute_2[62] = T.float32(0)
+ compute_2[63] = T.float32(0)
+ compute_2[64] = T.float32(0)
+ compute_2[65] = T.float32(0)
+ compute_2[66] = T.float32(0)
+ compute_2[67] = T.float32(0)
+ compute_2[68] = T.float32(0)
+ compute_2[69] = T.float32(0)
+ compute_2[70] = T.float32(0)
+ compute_2[71] = T.float32(0)
+ compute_2[72] = T.float32(0)
+ compute_2[73] = T.float32(0)
+ compute_2[74] = T.float32(0)
+ compute_2[75] = T.float32(0)
+ compute_2[76] = T.float32(0)
+ compute_2[77] = T.float32(0)
+ compute_2[78] = T.float32(0)
+ compute_2[79] = T.float32(0)
+ compute_2[80] = T.float32(0)
+ compute_2[81] = T.float32(0)
+ compute_2[82] = T.float32(0)
+ compute_2[83] = T.float32(0)
+ compute_2[84] = T.float32(0)
+ compute_2[85] = T.float32(0)
+ compute_2[86] = T.float32(0)
+ compute_2[87] = T.float32(0)
+ compute_2[88] = T.float32(0)
+ compute_2[89] = T.float32(0)
+ compute_2[90] = T.float32(0)
+ compute_2[91] = T.float32(0)
+ compute_2[92] = T.float32(0)
+ compute_2[93] = T.float32(0)
+ compute_2[94] = T.float32(0)
+ compute_2[95] = T.float32(0)
+ compute_2[96] = T.float32(0)
+ compute_2[97] = T.float32(0)
+ compute_2[98] = T.float32(0)
+ compute_2[99] = T.float32(0)
+ compute_2[100] = T.float32(0)
+ compute_2[101] = T.float32(0)
+ compute_2[102] = T.float32(0)
+ compute_2[103] = T.float32(0)
+ compute_2[104] = T.float32(0)
+ compute_2[105] = T.float32(0)
+ compute_2[106] = T.float32(0)
+ compute_2[107] = T.float32(0)
+ compute_2[108] = T.float32(0)
+ compute_2[109] = T.float32(0)
+ compute_2[110] = T.float32(0)
+ compute_2[111] = T.float32(0)
+ compute_2[112] = T.float32(0)
+ compute_2[113] = T.float32(0)
+ compute_2[114] = T.float32(0)
+ compute_2[115] = T.float32(0)
+ compute_2[116] = T.float32(0)
+ compute_2[117] = T.float32(0)
+ compute_2[118] = T.float32(0)
+ compute_2[119] = T.float32(0)
+ compute_2[120] = T.float32(0)
+ compute_2[121] = T.float32(0)
+ compute_2[122] = T.float32(0)
+ compute_2[123] = T.float32(0)
+ compute_2[124] = T.float32(0)
+ compute_2[125] = T.float32(0)
+ compute_2[126] = T.float32(0)
+ compute_2[127] = T.float32(0)
+ for elem_idx in range(placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
+ 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)
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[0] = compute_2[0] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[1] = compute_2[1] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[2] = compute_2[2] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[3] = compute_2[3] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[4] = compute_2[4] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[5] = compute_2[5] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[6] = compute_2[6] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[7] = compute_2[7] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[8] = compute_2[8] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[9] = compute_2[9] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[10] = compute_2[10] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[11] = compute_2[11] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[12] = compute_2[12] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[13] = compute_2[13] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[14] = compute_2[14] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[15] = compute_2[15] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx]], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[16] = compute_2[16] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[17] = compute_2[17] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[18] = compute_2[18] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[19] = compute_2[19] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[20] = compute_2[20] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[21] = compute_2[21] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[22] = compute_2[22] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[23] = compute_2[23] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[24] = compute_2[24] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[25] = compute_2[25] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[26] = compute_2[26] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[27] = compute_2[27] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[28] = compute_2[28] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[29] = compute_2[29] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[30] = compute_2[30] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[31] = compute_2[31] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 256], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[32] = compute_2[32] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[33] = compute_2[33] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[34] = compute_2[34] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[35] = compute_2[35] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[36] = compute_2[36] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[37] = compute_2[37] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[38] = compute_2[38] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[39] = compute_2[39] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[40] = compute_2[40] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[41] = compute_2[41] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[42] = compute_2[42] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[43] = compute_2[43] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[44] = compute_2[44] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[45] = compute_2[45] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[46] = compute_2[46] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[47] = compute_2[47] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 512], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[48] = compute_2[48] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[49] = compute_2[49] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[50] = compute_2[50] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[51] = compute_2[51] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[52] = compute_2[52] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[53] = compute_2[53] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[54] = compute_2[54] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[55] = compute_2[55] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[56] = compute_2[56] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[57] = compute_2[57] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[58] = compute_2[58] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[59] = compute_2[59] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[60] = compute_2[60] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[61] = compute_2[61] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[62] = compute_2[62] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[63] = compute_2[63] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 768], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[64] = compute_2[64] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[65] = compute_2[65] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[66] = compute_2[66] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[67] = compute_2[67] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[68] = compute_2[68] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[69] = compute_2[69] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[70] = compute_2[70] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[71] = compute_2[71] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[72] = compute_2[72] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[73] = compute_2[73] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[74] = compute_2[74] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[75] = compute_2[75] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[76] = compute_2[76] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[77] = compute_2[77] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[78] = compute_2[78] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[79] = compute_2[79] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1024], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[80] = compute_2[80] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[81] = compute_2[81] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[82] = compute_2[82] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[83] = compute_2[83] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[84] = compute_2[84] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[85] = compute_2[85] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[86] = compute_2[86] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[87] = compute_2[87] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[88] = compute_2[88] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[89] = compute_2[89] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[90] = compute_2[90] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[91] = compute_2[91] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[92] = compute_2[92] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[93] = compute_2[93] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[94] = compute_2[94] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[95] = compute_2[95] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1280], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[96] = compute_2[96] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[97] = compute_2[97] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[98] = compute_2[98] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[99] = compute_2[99] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[100] = compute_2[100] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[101] = compute_2[101] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[102] = compute_2[102] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[103] = compute_2[103] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[104] = compute_2[104] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[105] = compute_2[105] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[106] = compute_2[106] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[107] = compute_2[107] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[108] = compute_2[108] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[109] = compute_2[109] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[110] = compute_2[110] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[111] = compute_2[111] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1536], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[112] = compute_2[112] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[113] = compute_2[113] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 1] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[114] = compute_2[114] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 2] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[115] = compute_2[115] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 3] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[116] = compute_2[116] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 4] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[117] = compute_2[117] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 5] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[118] = compute_2[118] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 6] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[119] = compute_2[119] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 7] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[120] = compute_2[120] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 8] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[121] = compute_2[121] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 9] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[122] = compute_2[122] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 10] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[123] = compute_2[123] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 11] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[124] = compute_2[124] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 12] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[125] = compute_2[125] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 13] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[126] = compute_2[126] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 14] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ if T.likely(elem_idx < placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]):
+ compute_2[127] = compute_2[127] + placeholder_6[placeholder_5[cse_var_1] * 16 + elem_idx * 16 + 15] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + placeholder_8[placeholder_5[cse_var_1] + elem_idx] + 1792], T.float32(0))
+ for i0_inner in range(8):
+ cse_var_2: T.int32 = i0_outer_i1_outer_fused // 32 * 4096 + i0_inner * 512 + cse_var_1 * 16
+ compute_3 = T.Buffer((65536,), data=compute.data)
+ placeholder_5 = T.Buffer((65536,), data=placeholder_4.data)
+ compute_3[cse_var_2:cse_var_2 + 16] = T.max(compute_2[i0_inner * 16:i0_inner * 16 + 16] + placeholder_5[cse_var_2:cse_var_2 + 16], T.Broadcast(T.float32(0), 16))
</pre></div>
</div>
</div>
@@ -746,7 +1063,7 @@ 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: 1.814 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 2.847 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 83c9cb1b58..6691fef7fe 100644
--- a/docs/how_to/tune_with_autotvm/sg_execution_times.html
+++ b/docs/how_to/tune_with_autotvm/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-tune-with-autotvm-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:34.993</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
+<p><strong>00:38.201</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -349,22 +349,22 @@
</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>00:34.960</p></td>
+<td><p>00:38.168</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="tune_relay_x86.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-x86-py"><span class="std std-ref">Auto-tuning a Convolutional Network for x86 CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_x86.py</span></code>)</p></td>
-<td><p>00:00.019</p></td>
+<td><p>00:00.020</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="tune_relay_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-cuda-py"><span class="std std-ref">Auto-tuning a Convolutional Network for NVIDIA GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_cuda.py</span></code>)</p></td>
<td><p>00:00.005</p></td>
<td><p>0.0 MB</p></td>
</tr>
-<tr class="row-even"><td><p><a class="reference internal" href="tune_relay_arm.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-arm-py"><span class="std std-ref">Auto-tuning a Convolutional Network for ARM CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_arm.py</span></code>)</p></td>
+<tr class="row-even"><td><p><a class="reference internal" href="tune_relay_mobile_gpu.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-mobile-gpu-py"><span class="std std-ref">Auto-tuning a Convolutional Network for Mobile GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_mobile_gpu.py</span></code>)</p></td>
<td><p>00:00.004</p></td>
<td><p>0.0 MB</p></td>
</tr>
-<tr class="row-odd"><td><p><a class="reference internal" href="tune_relay_mobile_gpu.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-mobile-gpu-py"><span class="std std-ref">Auto-tuning a Convolutional Network for Mobile GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_mobile_gpu.py</span></code>)</p></td>
+<tr class="row-odd"><td><p><a class="reference internal" href="tune_relay_arm.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-arm-py"><span class="std std-ref">Auto-tuning a Convolutional Network for ARM CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_arm.py</span></code>)</p></td>
<td><p>00:00.004</p></td>
<td><p>0.0 MB</p></td>
</tr>
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 f04f54330c..479651078f 100644
--- a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
+++ b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
@@ -690,7 +690,7 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 64, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3173634
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 1, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3233179
No: 2 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
@@ -813,7 +813,7 @@ 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, 4, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1007576
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 64]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,890541
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)
@@ -936,10 +936,9 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 2, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6952174
-No: 4 GFLOPS: 61.52/61.52 result: MeasureResult(costs=(0.003763109814814815,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.1692252159118652, timestamp=1675844779.5716734) [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8912925
-No: 5 GFLOPS: 13.13/61.52 result: MeasureResult(costs=(0.01763057566666667,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.6716628074645996, timestamp=1675844783.3626807) [('tile_f', [-1, 1, 2, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7463608
-No: 6 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 8, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1372349
+No: 4 GFLOPS: 6.98/6.98 result: MeasureResult(costs=(0.03316484825,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.683012962341309, timestamp=1675876164.6765957) [('tile_f', [-1, 4, 2, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7559592
+No: 5 GFLOPS: 0.00/6.98 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1061,8 +1060,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 16, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6763057
-No: 7 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 256, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5485083
+No: 6 GFLOPS: 0.00/6.98 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
@@ -1184,8 +1183,8 @@ 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, 4, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1130603
-No: 8 GFLOPS: 0.00/61.52 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, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2731221
+No: 7 GFLOPS: 0.00/6.98 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
@@ -1307,9 +1306,9 @@ 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, 16, 8, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9282365
-No: 9 GFLOPS: 8.43/61.52 result: MeasureResult(costs=(0.027449855750000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.60951566696167, timestamp=1675844785.193633) [('tile_f', [-1, 2, 1, 64]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,431401
-No: 10 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 8, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 64, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10184759
+No: 8 GFLOPS: 394.41/394.41 result: MeasureResult(costs=(0.0005869529289099526,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4507114887237549, timestamp=1675876167.2906568) [('tile_f', [-1, 1, 8, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2237647
+No: 9 GFLOPS: 0.00/394.41 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1431,8 +1430,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 2, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7462074
-No: 11 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 4, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 64, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6118936
+No: 10 GFLOPS: 0.00/394.41 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
@@ -1554,8 +1553,9 @@ 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, 16, 4, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 512]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1160179
-No: 12 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 256, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,6985273
+No: 11 GFLOPS: 44.86/394.41 result: MeasureResult(costs=(0.00516085775,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.5173752307891846, timestamp=1675876168.9924073) [('tile_f', [-1, 2, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5079401
+No: 12 GFLOPS: 0.00/394.41 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
@@ -1677,8 +1677,8 @@ 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, 1, 256]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5088157
-No: 13 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 256, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2353351
+No: 13 GFLOPS: 0.00/394.41 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
@@ -1800,8 +1800,9 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 4, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5505651
-No: 14 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+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, 7, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7696988
+No: 14 GFLOPS: 8.30/394.41 result: MeasureResult(costs=(0.0279019525,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8869495391845703, timestamp=1675876174.5089574) [('tile_f', [-1, 4, 8, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6004981
+No: 15 GFLOPS: 0.00/394.41 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
@@ -1923,8 +1924,9 @@ 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, 16, 8, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8404083
-No: 15 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 16, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 128, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2964846
+No: 16 GFLOPS: 485.98/485.98 result: MeasureResult(costs=(0.0004763566886227545,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.412088871002197, timestamp=1675876175.5081959) [('tile_f', [-1, 1, 16, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8988134
+No: 17 GFLOPS: 0.00/485.98 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
@@ -2046,8 +2048,8 @@ 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, 1, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7376001
-No: 16 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 2, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,358584
+No: 18 GFLOPS: 0.00/485.98 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
@@ -2169,8 +2171,8 @@ 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, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 256, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6066597
-No: 17 GFLOPS: 0.00/61.52 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 64, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2802054
+No: 19 GFLOPS: 0.00/485.98 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
@@ -2292,284 +2294,8 @@ 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, 4, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4585536
-No: 18 GFLOPS: 2.23/61.52 result: MeasureResult(costs=(0.10378480525,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.782615900039673, timestamp=1675844790.4885643) [('tile_f', [-1, 16, 1, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6649999
-No: 19 GFLOPS: 0.00/61.52 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 357, 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 171, 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: 0x00007fabf2bc9fa2
- 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:185
- 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:1617
- 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
- at ../include/tvm/runtime/packed_func.h:1217
- 1: Call
- at ../include/tvm/runtime/packed_func.h:1213
- 0: operator()
- at ../src/runtime/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, 1, 1, 512]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9877339
-No: 20 GFLOPS: 0.00/61.52 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:1730
- 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
- at ../include/tvm/runtime/packed_func.h:1670
- 19: run<>
- at ../include/tvm/runtime/packed_func.h:1630
- 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1645
- 13: operator()
- at ../src/driver/driver_api.cc:395
- 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
- at ../src/driver/driver_api.cc:381
- 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
- at ../src/driver/driver_api.cc:276
- 10: tvm::transform::Pass::operator()(tvm::IRModule) const
- at ../src/ir/transform.cc:258
- 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:274
- 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:451
- 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:274
- 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/tir/ir/transform.cc:100
- 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
- at ../include/tvm/runtime/packed_func.h:1749
- 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
- at ../include/tvm/runtime/packed_func.h:1693
- 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
- at ../include/tvm/runtime/packed_func.h:1617
- 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
- at ../include/tvm/runtime/packed_func.h:1217
- 1: Call
- at ../include/tvm/runtime/packed_func.h:1213
- 0: operator()
- at ../src/runtime/c_runtime_api.cc:534
- File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
- File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
- raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel
-
-Traceback (most recent call last):
- 24: TVMFuncCall
- at ../src/runtime/c_runtime_api.cc:477
- 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
- at ../include/tvm/runtime/packed_func.h:1217
- 22: Call
- at ../include/tvm/runtime/packed_func.h:1213
- 21: operator()
- at ../include/tvm/runtime/packed_func.h:1730
- 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
- at ../include/tvm/runtime/packed_func.h:1670
- 19: run<>
- at ../include/tvm/runtime/packed_func.h:1630
- 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1630
- 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
- at ../include/tvm/runtime/packed_func.h:1645
- 13: operator()
- at ../src/driver/driver_api.cc:395
- 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
- at ../src/driver/driver_api.cc:381
- 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
- at ../src/driver/driver_api.cc:276
- 10: tvm::transform::Pass::operator()(tvm::IRModule) const
- at ../src/ir/transform.cc:258
- 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:274
- 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:451
- 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/ir/transform.cc:274
- 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
- at ../src/tir/ir/transform.cc:100
- 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
- at ../include/tvm/runtime/packed_func.h:1749
- 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
- at ../include/tvm/runtime/packed_func.h:1693
- 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
- at ../include/tvm/runtime/packed_func.h:1617
- 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
- at ../include/tvm/runtime/packed_func.h:1217
- 1: Call
- at ../include/tvm/runtime/packed_func.h:1213
- 0: operator()
- at ../src/runtime/c_runtime_api.cc:534
- File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
- File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
- raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 1, 2]), ('tile_y', [-1, 7, 1, 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', 512), ('unroll_explicit', 0)],None,2257481
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 32, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 8, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2016564
+No: 20 GFLOPS: 253.61/485.98 result: MeasureResult(costs=(0.0009128193448275862,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.485121250152588, timestamp=1675876177.2099645) [('tile_f', [-1, 1, 8, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2419021
</pre></div>
</div>
<p>Finally we can inspect the best config from log file, check correctness,
@@ -2608,9 +2334,9 @@ 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, 2, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8912925
+[('tile_f', [-1, 1, 16, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8988134
Finish loading 20 records
-Time cost of this operator: 0.001784
+Time cost of this operator: 0.000840
</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 0b7db79050..adcbdd0b40 100644
--- a/docs/how_to/work_with_microtvm/micro_autotune.html
+++ b/docs/how_to/work_with_microtvm/micro_autotune.html
@@ -643,10 +643,10 @@ 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 311.2 98.689 (1, 2, 10, 10, 3) 2 1 [311.2]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.125 0.991 (1, 6, 10, 10) 1 1 [3.125]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 1.01 0.32 (1, 1, 10, 10, 3) 1 1 [1.01]
-Total_time - 315.335 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 308.5 98.68 (1, 2, 10, 10, 3) 2 1 [308.5]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.164 1.012 (1, 6, 10, 10) 1 1 [3.164]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.963 0.308 (1, 1, 10, 10, 3) 1 1 [0.963]
+Total_time - 312.627 - - - - -
</pre></div>
</div>
</div>
@@ -698,10 +698,10 @@ Total_time -
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>########## Build with Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs Measurements(us)
--------- --- -------- ------- ----- ------ ------- ----------------
-tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 103.2 97.508 (1, 6, 10, 10, 1) 2 1 [103.2]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.794 1.695 (1, 6, 10, 10) 1 1 [1.794]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.843 0.797 (1, 3, 10, 10, 1) 1 1 [0.843]
-Total_time - 105.837 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 100.2 97.295 (1, 6, 10, 10, 1) 2 1 [100.2]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.816 1.763 (1, 6, 10, 10) 1 1 [1.816]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.97 0.942 (1, 1, 10, 10, 3) 1 1 [0.97]
+Total_time - 102.986 - - - - -
</pre></div>
</div>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-work-with-microtvm-micro-autotune-py">
diff --git a/docs/how_to/work_with_microtvm/micro_pytorch.html b/docs/how_to/work_with_microtvm/micro_pytorch.html
index 92ba77dbab..a3c199189f 100644
--- a/docs/how_to/work_with_microtvm/micro_pytorch.html
+++ b/docs/how_to/work_with_microtvm/micro_pytorch.html
@@ -454,7 +454,8 @@ 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, 43.7MB/s]
+ 61%|###### | 2.09M/3.42M [00:00<00:00, 13.0MB/s]
+100%|##########| 3.42M/3.42M [00:00<00:00, 20.4MB/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.
@@ -580,7 +581,7 @@ via the host <cite>main.cc`</cite> or if a Zephyr emulated board is selected as
Torch top-1 id: 282, class name: tiger cat
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 11.212 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 10.246 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 6ee274a556..caf57f9cd4 100644
--- a/docs/how_to/work_with_microtvm/micro_train.html
+++ b/docs/how_to/work_with_microtvm/micro_train.html
@@ -523,7 +523,7 @@ take about <strong>2 minutes</strong> to download the Stanford Cars, while COCO
<a href="https://docs.python.org/3/library/shutil.html#shutil.move" title="shutil.move" class="sphx-glr-backref-module-shutil sphx-glr-backref-type-py-function"><span class="n">shutil</span><span class="o">.</span><span class="n">move</span></a><span class="p">(</span><span class="sa">f</span><span class="s2">"</span><span class="si">{</span><a href="https://docs.python.org/3/library/stdtypes.html#str" title="builtins.str" class="sphx-glr-backref-module-builtins sphx-glr-backref-typ [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>'/tmp/tmpm50rqx90/images/random'
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>'/tmp/tmpy9cmecff/images/random'
</pre></div>
</div>
</div>
@@ -583,8 +583,8 @@ objects to other stuff? We can display some examples from our datasets using <co
<span class="n">plt</span><span class="o">.</span><span class="n">axis</span><span class="p">(</span><span class="s2">"off"</span><span class="p">)</span>
</pre></div>
</div>
-<img src="../../_images/sphx_glr_micro_train_001.png" srcset="../../_images/sphx_glr_micro_train_001.png" alt="[0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [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/tmpm50rqx90/images/target contains 8144 images
-/tmp/tmpm50rqx90/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/tmpy9cmecff/images/target contains 8144 images
+/tmp/tmpy9cmecff/images/random contains 5000 images
</pre></div>
</div>
</div>
@@ -696,13 +696,13 @@ the time on our validation set).</p>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Epoch 1/3
-328/328 - 47s - loss: 0.2678 - accuracy: 0.9141 - val_loss: 0.1673 - val_accuracy: 0.9479 - 47s/epoch - 142ms/step
+328/328 - 47s - loss: 0.2250 - accuracy: 0.9227 - val_loss: 0.1374 - val_accuracy: 0.9528 - 47s/epoch - 144ms/step
Epoch 2/3
-328/328 - 43s - loss: 0.1014 - accuracy: 0.9633 - val_loss: 0.1538 - val_accuracy: 0.9543 - 43s/epoch - 131ms/step
+328/328 - 44s - loss: 0.0961 - accuracy: 0.9655 - val_loss: 0.1376 - val_accuracy: 0.9569 - 44s/epoch - 133ms/step
Epoch 3/3
-328/328 - 43s - loss: 0.0677 - accuracy: 0.9747 - val_loss: 0.1398 - val_accuracy: 0.9611 - 43s/epoch - 131ms/step
+328/328 - 44s - loss: 0.0702 - accuracy: 0.9747 - val_loss: 0.1168 - val_accuracy: 0.9603 - 44s/epoch - 133ms/step
-<keras.callbacks.History object at 0x7f1c024aad90>
+<keras.callbacks.History object at 0x7efdc2500410>
</pre></div>
</div>
</div>
@@ -966,7 +966,7 @@ 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> ( 5 minutes 2.592 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 4 minutes 59.015 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 4275c8588c..d8ad3a77ea 100644
--- a/docs/how_to/work_with_microtvm/sg_execution_times.html
+++ b/docs/how_to/work_with_microtvm/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-work-with-microtvm-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>07:19.778</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
+<p><strong>07:14.192</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 82%" />
@@ -349,23 +349,23 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="micro_train.html#sphx-glr-how-to-work-with-microtvm-micro-train-py"><span class="std std-ref">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>05:02.592</p></td>
+<td><p>04:59.015</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="micro_pytorch.html#sphx-glr-how-to-work-with-microtvm-micro-pytorch-py"><span class="std std-ref">4. microTVM PyTorch Tutorial</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_pytorch.py</span></code>)</p></td>
-<td><p>01:11.212</p></td>
+<td><p>01:10.246</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="micro_autotune.html#sphx-glr-how-to-work-with-microtvm-micro-autotune-py"><span class="std std-ref">6. Model Tuning with microTVM</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_autotune.py</span></code>)</p></td>
-<td><p>00:52.655</p></td>
+<td><p>00:52.052</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>00:08.042</p></td>
+<td><p>00:07.823</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="micro_tflite.html#sphx-glr-how-to-work-with-microtvm-micro-tflite-py"><span class="std std-ref">2. microTVM TFLite Tutorial</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_tflite.py</span></code>)</p></td>
-<td><p>00:05.278</p></td>
+<td><p>00:05.055</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="micro_ethosu.html#sphx-glr-how-to-work-with-microtvm-micro-ethosu-py"><span class="std std-ref">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 4967b2c5b9..c327451061 100644
--- a/docs/how_to/work_with_relay/sg_execution_times.html
+++ b/docs/how_to/work_with_relay/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-work-with-relay-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:44.345</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
+<p><strong>00:44.387</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -349,15 +349,15 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="using_pipeline_executor.html#sphx-glr-how-to-work-with-relay-using-pipeline-executor-py"><span class="std std-ref">Using Pipeline Executor in Relay</span></a> (<code class="docutils literal notranslate"><span class="pre">using_pipeline_executor.py</span></code>)</p></td>
-<td><p>00:32.208</p></td>
+<td><p>00:32.249</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="using_external_lib.html#sphx-glr-how-to-work-with-relay-using-external-lib-py"><span class="std std-ref">Using External Libraries in Relay</span></a> (<code class="docutils literal notranslate"><span class="pre">using_external_lib.py</span></code>)</p></td>
-<td><p>00:10.584</p></td>
+<td><p>00:10.435</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>00:01.547</p></td>
+<td><p>00:01.697</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 f0694452a1..aa270f8c0f 100644
--- a/docs/how_to/work_with_schedules/intrin_math.html
+++ b/docs/how_to/work_with_schedules/intrin_math.html
@@ -535,7 +535,7 @@ The following example customizes CUDA lowering rule for <code class="code docuti
<a href="../../reference/api/python/ir.html#tvm.ir.register_intrin_lowering" title="tvm.ir.register_intrin_lowering" class="sphx-glr-backref-module-tvm-ir sphx-glr-backref-type-py-function"><span class="n">register_intrin_lowering</span></a><span class="p">(</span><span class="s2">"tir.exp"</span><span class="p">,</span> <span class="n">target</span><span class="o">=</span><span class="s2">"cuda"</span><span class="p">,</span> <span class="n">f</span><span class="o">= [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span><function my_cuda_math_rule at 0x7f1aaf1514d0>
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span><function my_cuda_math_rule at 0x7efc6c911560>
</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 212156687b..091da5d131 100644
--- a/docs/how_to/work_with_schedules/sg_execution_times.html
+++ b/docs/how_to/work_with_schedules/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-work-with-schedules-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:06.622</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
+<p><strong>00:08.952</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -349,23 +349,23 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="intrin_math.html#sphx-glr-how-to-work-with-schedules-intrin-math-py"><span class="std std-ref">Intrinsics and Math Functions</span></a> (<code class="docutils literal notranslate"><span class="pre">intrin_math.py</span></code>)</p></td>
-<td><p>00:04.061</p></td>
+<td><p>00:06.420</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>00:01.195</p></td>
+<td><p>00:01.177</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>00:00.576</p></td>
+<td><p>00:00.577</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>00:00.559</p></td>
+<td><p>00:00.552</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="extern_op.html#sphx-glr-how-to-work-with-schedules-extern-op-py"><span class="std std-ref">External Tensor Functions</span></a> (<code class="docutils literal notranslate"><span class="pre">extern_op.py</span></code>)</p></td>
-<td><p>00:00.118</p></td>
+<td><p>00:00.115</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="schedule_primitives.html#sphx-glr-how-to-work-with-schedules-schedule-primitives-py"><span class="std std-ref">Schedule Primitives in TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">schedule_primitives.py</span></code>)</p></td>
@@ -373,7 +373,7 @@
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="tedd.html#sphx-glr-how-to-work-with-schedules-tedd-py"><span class="std std-ref">Use Tensor Expression Debug Display (TEDD) for Visualization</span></a> (<code class="docutils literal notranslate"><span class="pre">tedd.py</span></code>)</p></td>
-<td><p>00:00.033</p></td>
+<td><p>00:00.032</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/how_to/work_with_schedules/tensorize.html b/docs/how_to/work_with_schedules/tensorize.html
index 5a031f62b4..47f3d31ad7 100644
--- a/docs/how_to/work_with_schedules/tensorize.html
+++ b/docs/how_to/work_with_schedules/tensorize.html
@@ -586,7 +586,7 @@ class Module:
def main(A: T.Buffer((1024, 64), "float32"), B: T.Buffer((512, 64), "float32"), C: T.Buffer((1024, 512), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
i = T.var("int32")
- T.attr(T.iter_var(i, None, "DataPar", ""), "pragma_import_llvm", "; ModuleID = '/tmp/tmpzoz5ee0i/input0.cc'\nsource_filename = \"/tmp/tmpzoz5ee0i/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = alloca [...]
+ T.attr(T.iter_var(i, None, "DataPar", ""), "pragma_import_llvm", "; ModuleID = '/tmp/tmpxfbq4px4/input0.cc'\nsource_filename = \"/tmp/tmpxfbq4px4/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = alloca [...]
for i, j_outer in T.grid(1024, 32):
T.call_extern("int32", "gemv_update", T.tvm_access_ptr(T.type_annotation("float32"), C.data, i * 512 + j_outer * 16, 16, 2), T.tvm_access_ptr(T.type_annotation("float32"), A.data, i * 64, 64, 1), T.tvm_access_ptr(T.type_annotation("float32"), B.data, j_outer * 1024, 1024, 1), 16, 64, 64)
</pre></div>
diff --git a/docs/reference/api/python/auto_scheduler.html b/docs/reference/api/python/auto_scheduler.html
index f54524e134..ac9b1f12b0 100644
--- a/docs/reference/api/python/auto_scheduler.html
+++ b/docs/reference/api/python/auto_scheduler.html
@@ -1616,7 +1616,7 @@ 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>
@@ -1900,7 +1900,7 @@ 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 3845fce5bc..5473e742d3 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/c36ae1cb3/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -141,7 +141,7 @@
<div class="tsd-signature tsd-kind-icon">bytes<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Uint8Array</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
</ul>
</aside>
</section>
@@ -151,7 +151,7 @@
<div class="tsd-signature tsd-kind-icon">offset<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span><span class="tsd-signature-symbol"> = 0</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L42">rpc_server.ts:42</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L42">rpc_server.ts:42</a></li>
</ul>
</aside>
</section>
@@ -168,7 +168,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L63">rpc_server.ts:63</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L63">rpc_server.ts:63</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">Uint8Array</span></h4>
@@ -185,7 +185,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L49">rpc_server.ts:49</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L49">rpc_server.ts:49</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
@@ -202,7 +202,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L57">rpc_server.ts:57</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L57">rpc_server.ts:57</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
diff --git a/docs/reference/api/typedoc/classes/cachedcallstack.html b/docs/reference/api/typedoc/classes/cachedcallstack.html
index 6988627028..27c55737ef 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/c36ae1cb3/web/src/memory.ts#L223">memory.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L208">memory.ts:208</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L312">memory.ts:312</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L284">memory.ts:284</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L388">memory.ts:388</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L376">memory.ts:376</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L267">memory.ts:267</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L243">memory.ts:243</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L321">memory.ts:321</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L252">memory.ts:252</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L359">memory.ts:359</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L342">memory.ts:342</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L350">memory.ts:350</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L326">memory.ts:326</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L363">memory.ts:363</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L346">memory.ts:346</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/memory.ts#L334">memory.ts:334</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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 d21760f695..1385d55856 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/c36ae1cb3/web/src/runtime.ts#L262">runtime.ts:262</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L262">runtime.ts:262</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -147,7 +147,7 @@
<div class="tsd-signature tsd-kind-icon">bits<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L260">runtime.ts:260</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L260">runtime.ts:260</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -162,7 +162,7 @@
<div class="tsd-signature tsd-kind-icon">code<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L258">runtime.ts:258</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L258">runtime.ts:258</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -177,7 +177,7 @@
<div class="tsd-signature tsd-kind-icon">lanes<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L262">runtime.ts:262</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L262">runtime.ts:262</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -199,7 +199,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L279">runtime.ts:279</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L279">runtime.ts:279</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
@@ -216,7 +216,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L270">runtime.ts:270</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L270">runtime.ts:270</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">string</span></h4>
diff --git a/docs/reference/api/typedoc/classes/dldevice.html b/docs/reference/api/typedoc/classes/dldevice.html
index d3d95c7c4b..b0689f9d98 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/c36ae1cb3/web/src/runtime.ts#L202">runtime.ts:202</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L202">runtime.ts:202</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -146,7 +146,7 @@
<div class="tsd-signature tsd-kind-icon">device<wbr>Id<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L200">runtime.ts:200</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L200">runtime.ts:200</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -161,7 +161,7 @@
<div class="tsd-signature tsd-kind-icon">device<wbr>Type<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L198">runtime.ts:198</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L198">runtime.ts:198</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -183,7 +183,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L223">runtime.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L223">runtime.ts:223</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -205,7 +205,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L230">runtime.ts:230</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L230">runtime.ts:230</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">string</span></h4>
diff --git a/docs/reference/api/typedoc/classes/environment.html b/docs/reference/api/typedoc/classes/environment.html
index 350560babb..da68ffe97f 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/c36ae1cb3/web/src/environment.ts#L86">environment.ts:86</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/environment.ts#L70">environment.ts:70</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/environment.ts#L69">environment.ts:69</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/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/c36ae1cb3/web/src/environment.ts#L78">environment.ts:78</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/environment.ts#L78">environment.ts:78</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -228,7 +228,7 @@
<div class="tsd-signature tsd-kind-icon">packedCFunc<wbr>Table<wbr>Free<wbr>Id<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">></span><span class="tsd-signature-symbol"> = []</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/environment.ts#L84">environment.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/environment.ts#L84">environment.ts:84</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -250,7 +250,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/environment.ts#L105">environment.ts:105</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/environment.ts#L105">environment.ts:105</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/ffilibrary.html b/docs/reference/api/typedoc/classes/ffilibrary.html
index 8e36e5dd8a..7398a93e3e 100644
--- a/docs/reference/api/typedoc/classes/ffilibrary.html
+++ b/docs/reference/api/typedoc/classes/ffilibrary.html
@@ -131,7 +131,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L49">runtime.ts:49</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L49">runtime.ts:49</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -156,7 +156,7 @@
<div class="tsd-signature tsd-kind-icon">exports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">Function</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L46">runtime.ts:46</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L46">runtime.ts:46</a></li>
</ul>
</aside>
</section>
@@ -166,7 +166,7 @@
<div class="tsd-signature tsd-kind-icon">memory<span class="tsd-signature-symbol">:</span> <a href="memory.html" class="tsd-signature-type">Memory</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L45">runtime.ts:45</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L45">runtime.ts:45</a></li>
</ul>
</aside>
</section>
@@ -176,7 +176,7 @@
<div class="tsd-signature tsd-kind-icon">wasm32<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">boolean</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L44">runtime.ts:44</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L44">runtime.ts:44</a></li>
</ul>
</aside>
</section>
@@ -186,7 +186,7 @@
<div class="tsd-signature tsd-kind-icon">webGPUContext<span class="tsd-signature-symbol">:</span> <a href="webgpucontext.html" class="tsd-signature-type">WebGPUContext</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L47">runtime.ts:47</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L47">runtime.ts:47</a></li>
</ul>
</aside>
</section>
@@ -203,7 +203,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L76">runtime.ts:76</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L76">runtime.ts:76</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -226,7 +226,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L66">runtime.ts:66</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L66">runtime.ts:66</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -243,7 +243,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L84">runtime.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L84">runtime.ts:84</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <a href="cachedcallstack.html" class="tsd-signature-type">CachedCallStack</a></h4>
@@ -260,7 +260,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L95">runtime.ts:95</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L95">runtime.ts:95</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -283,7 +283,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L72">runtime.ts:72</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L72">runtime.ts:72</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
diff --git a/docs/reference/api/typedoc/classes/graphexecutor.html b/docs/reference/api/typedoc/classes/graphexecutor.html
index b2be6f8bde..4446cbcf2b 100644
--- a/docs/reference/api/typedoc/classes/graphexecutor.html
+++ b/docs/reference/api/typedoc/classes/graphexecutor.html
@@ -130,7 +130,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L583">runtime.ts:583</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L583">runtime.ts:583</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -162,7 +162,7 @@
<div class="tsd-signature tsd-kind-icon">module<span class="tsd-signature-symbol">:</span> <a href="module.html" class="tsd-signature-type">Module</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L579">runtime.ts:579</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L579">runtime.ts:579</a></li>
</ul>
</aside>
</section>
@@ -179,7 +179,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L654">runtime.ts:654</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L654">runtime.ts:654</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -224,7 +224,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L597">runtime.ts:597</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L597">runtime.ts:597</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -241,7 +241,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L631">runtime.ts:631</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L631">runtime.ts:631</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -279,7 +279,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L644">runtime.ts:644</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L644">runtime.ts:644</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -310,7 +310,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L621">runtime.ts:621</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L621">runtime.ts:621</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -332,7 +332,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L609">runtime.ts:609</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L609">runtime.ts:609</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/instance.html b/docs/reference/api/typedoc/classes/instance.html
index a11017d592..a38caf7133 100644
--- a/docs/reference/api/typedoc/classes/instance.html
+++ b/docs/reference/api/typedoc/classes/instance.html
@@ -139,7 +139,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L692">runtime.ts:692</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L692">runtime.ts:692</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -202,7 +202,7 @@
<div class="tsd-signature tsd-kind-icon">exports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">Function</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L684">runtime.ts:684</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L684">runtime.ts:684</a></li>
</ul>
</aside>
</section>
@@ -212,7 +212,7 @@
<div class="tsd-signature tsd-kind-icon">memory<span class="tsd-signature-symbol">:</span> <a href="memory.html" class="tsd-signature-type">Memory</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L683">runtime.ts:683</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L683">runtime.ts:683</a></li>
</ul>
</aside>
</section>
@@ -229,7 +229,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L932">runtime.ts:932</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L932">runtime.ts:932</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -260,7 +260,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L994">runtime.ts:994</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L994">runtime.ts:994</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -303,7 +303,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L924">runtime.ts:924</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L924">runtime.ts:924</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -341,7 +341,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L732">runtime.ts:732</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L732">runtime.ts:732</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -358,7 +358,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L952">runtime.ts:952</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L952">runtime.ts:952</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -402,7 +402,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L816">runtime.ts:816</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L816">runtime.ts:816</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -434,7 +434,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L1033">runtime.ts:1033</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L1033">runtime.ts:1033</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -465,7 +465,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L846">runtime.ts:846</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L846">runtime.ts:846</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -497,7 +497,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L750">runtime.ts:750</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L750">runtime.ts:750</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -520,7 +520,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L1013">runtime.ts:1013</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L1013">runtime.ts:1013</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -568,7 +568,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L789">runtime.ts:789</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L789">runtime.ts:789</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -608,7 +608,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L914">runtime.ts:914</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L914">runtime.ts:914</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -646,7 +646,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L1145">runtime.ts:1145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L1145">runtime.ts:1145</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -698,7 +698,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L740">runtime.ts:740</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L740">runtime.ts:740</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -722,7 +722,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L868">runtime.ts:868</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L868">runtime.ts:868</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -754,7 +754,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L857">runtime.ts:857</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L857">runtime.ts:857</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -786,7 +786,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L940">runtime.ts:940</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L940">runtime.ts:940</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/memory.html b/docs/reference/api/typedoc/classes/memory.html
index d3b42ce71e..2c44a6fe11 100644
--- a/docs/reference/api/typedoc/classes/memory.html
+++ b/docs/reference/api/typedoc/classes/memory.html
@@ -130,7 +130,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L40">memory.ts:40</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L40">memory.ts:40</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -152,7 +152,7 @@
<div class="tsd-signature tsd-kind-icon">memory<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Memory</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L32">memory.ts:32</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L32">memory.ts:32</a></li>
</ul>
</aside>
</section>
@@ -162,7 +162,7 @@
<div class="tsd-signature tsd-kind-icon">wasm32<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">boolean</span><span class="tsd-signature-symbol"> = true</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L33">memory.ts:33</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L33">memory.ts:33</a></li>
</ul>
</aside>
</section>
@@ -179,7 +179,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L154">memory.ts:154</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L154">memory.ts:154</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -210,7 +210,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L90">memory.ts:90</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L90">memory.ts:90</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -233,7 +233,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L97">memory.ts:97</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L97">memory.ts:97</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -256,7 +256,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L74">memory.ts:74</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L74">memory.ts:74</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -279,7 +279,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L81">memory.ts:81</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L81">memory.ts:81</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -302,7 +302,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L104">memory.ts:104</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L104">memory.ts:104</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -325,7 +325,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L132">memory.ts:132</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L132">memory.ts:132</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -362,7 +362,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L145">memory.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L145">memory.ts:145</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -393,7 +393,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L60">memory.ts:60</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L60">memory.ts:60</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -416,7 +416,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L67">memory.ts:67</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L67">memory.ts:67</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -439,7 +439,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L53">memory.ts:53</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L53">memory.ts:53</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -462,7 +462,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L114">memory.ts:114</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L114">memory.ts:114</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -485,7 +485,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L124">memory.ts:124</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L124">memory.ts:124</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
@@ -502,7 +502,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/memory.ts#L175">memory.ts:175</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/memory.ts#L175">memory.ts:175</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/module.html b/docs/reference/api/typedoc/classes/module.html
index dc5bd383b2..ae7c25becd 100644
--- a/docs/reference/api/typedoc/classes/module.html
+++ b/docs/reference/api/typedoc/classes/module.html
@@ -124,7 +124,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L504">runtime.ts:504</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L504">runtime.ts:504</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -170,7 +170,7 @@
<div class="tsd-signature tsd-kind-icon">handle<span class="tsd-signature-symbol">:</span> <a href="../index.html#pointer" class="tsd-signature-type">Pointer</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L502">runtime.ts:502</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L502">runtime.ts:502</a></li>
</ul>
</aside>
</section>
@@ -187,7 +187,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L516">runtime.ts:516</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L516">runtime.ts:516</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -204,7 +204,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L530">runtime.ts:530</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L530">runtime.ts:530</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -236,7 +236,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L561">runtime.ts:561</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L561">runtime.ts:561</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/ndarray.html b/docs/reference/api/typedoc/classes/ndarray.html
index 5aee3e746e..9652f7ad38 100644
--- a/docs/reference/api/typedoc/classes/ndarray.html
+++ b/docs/reference/api/typedoc/classes/ndarray.html
@@ -130,7 +130,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L304">runtime.ts:304</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L304">runtime.ts:304</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -158,7 +158,7 @@
<div class="tsd-signature tsd-kind-icon">device<span class="tsd-signature-symbol">:</span> <a href="dldevice.html" class="tsd-signature-type">DLDevice</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L297">runtime.ts:297</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L297">runtime.ts:297</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -173,7 +173,7 @@
<div class="tsd-signature tsd-kind-icon">dtype<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">string</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L293">runtime.ts:293</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L293">runtime.ts:293</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -188,7 +188,7 @@
<div class="tsd-signature tsd-kind-icon">handle<span class="tsd-signature-symbol">:</span> <a href="../index.html#pointer" class="tsd-signature-type">Pointer</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L289">runtime.ts:289</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L289">runtime.ts:289</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -203,7 +203,7 @@
<div class="tsd-signature tsd-kind-icon">ndim<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L291">runtime.ts:291</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L291">runtime.ts:291</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -218,7 +218,7 @@
<div class="tsd-signature tsd-kind-icon">shape<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L295">runtime.ts:295</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L295">runtime.ts:295</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -240,7 +240,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L370">runtime.ts:370</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L370">runtime.ts:370</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -273,7 +273,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L414">runtime.ts:414</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L414">runtime.ts:414</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -305,7 +305,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L355">runtime.ts:355</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L355">runtime.ts:355</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -322,7 +322,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L474">runtime.ts:474</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L474">runtime.ts:474</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -346,7 +346,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L443">runtime.ts:443</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L443">runtime.ts:443</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/packedfunccell.html b/docs/reference/api/typedoc/classes/packedfunccell.html
index 1c935173c2..c4047a3cf6 100644
--- a/docs/reference/api/typedoc/classes/packedfunccell.html
+++ b/docs/reference/api/typedoc/classes/packedfunccell.html
@@ -122,7 +122,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L158">runtime.ts:158</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L158">runtime.ts:158</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -147,7 +147,7 @@
<div class="tsd-signature tsd-kind-icon">handle<span class="tsd-signature-symbol">:</span> <a href="../index.html#pointer" class="tsd-signature-type">Pointer</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L157">runtime.ts:157</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L157">runtime.ts:157</a></li>
</ul>
</aside>
</section>
@@ -164,7 +164,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L165">runtime.ts:165</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L165">runtime.ts:165</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
diff --git a/docs/reference/api/typedoc/classes/rpcserver.html b/docs/reference/api/typedoc/classes/rpcserver.html
index 93b9a940b6..e0206909ee 100644
--- a/docs/reference/api/typedoc/classes/rpcserver.html
+++ b/docs/reference/api/typedoc/classes/rpcserver.html
@@ -115,7 +115,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L92">rpc_server.ts:92</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L92">rpc_server.ts:92</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -176,7 +176,7 @@
<div class="tsd-signature tsd-kind-icon">get<wbr>Imports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> => </span><span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">unknown</span><span class="tsd-signat [...]
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L82">rpc_server.ts:82</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L82">rpc_server.ts:82</a></li>
</ul>
</aside>
<div class="tsd-type-declaration">
@@ -201,7 +201,7 @@
<div class="tsd-signature tsd-kind-icon">key<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">string</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L78">rpc_server.ts:78</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L78">rpc_server.ts:78</a></li>
</ul>
</aside>
</section>
@@ -211,7 +211,7 @@
<div class="tsd-signature tsd-kind-icon">logger<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>msg<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> => </span><span class="tsd-signature-type">void</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L81">rpc_server.ts:81</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L81">rpc_server.ts:81</a></li>
</ul>
</aside>
<div class="tsd-type-declaration">
@@ -242,7 +242,7 @@
<div class="tsd-signature tsd-kind-icon">socket<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">WebSocket</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L79">rpc_server.ts:79</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L79">rpc_server.ts:79</a></li>
</ul>
</aside>
</section>
@@ -252,7 +252,7 @@
<div class="tsd-signature tsd-kind-icon">state<span class="tsd-signature-symbol">:</span> <a href="../enums/rpcserverstate.html" class="tsd-signature-type">RPCServerState</a><span class="tsd-signature-symbol"> = RPCServerState.InitHeader</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L80">rpc_server.ts:80</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L80">rpc_server.ts:80</a></li>
</ul>
</aside>
</section>
@@ -262,7 +262,7 @@
<div class="tsd-signature tsd-kind-icon">url<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">string</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/rpc_server.ts#L77">rpc_server.ts:77</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/rpc_server.ts#L77">rpc_server.ts:77</a></li>
</ul>
</aside>
</section>
diff --git a/docs/reference/api/typedoc/classes/scalar.html b/docs/reference/api/typedoc/classes/scalar.html
index edff28a85c..541507dd48 100644
--- a/docs/reference/api/typedoc/classes/scalar.html
+++ b/docs/reference/api/typedoc/classes/scalar.html
@@ -112,7 +112,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L145">runtime.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L145">runtime.ts:145</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -137,7 +137,7 @@
<div class="tsd-signature tsd-kind-icon">dtype<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">string</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L145">runtime.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L145">runtime.ts:145</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -152,7 +152,7 @@
<div class="tsd-signature tsd-kind-icon">value<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/runtime.ts#L143">runtime.ts:143</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/runtime.ts#L143">runtime.ts:143</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/webgpucontext.html b/docs/reference/api/typedoc/classes/webgpucontext.html
index f9afe257a3..c4aa4d52cb 100644
--- a/docs/reference/api/typedoc/classes/webgpucontext.html
+++ b/docs/reference/api/typedoc/classes/webgpucontext.html
@@ -120,7 +120,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/webgpu.ts#L57">webgpu.ts:57</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/webgpu.ts#L57">webgpu.ts:57</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -145,7 +145,7 @@
<div class="tsd-signature tsd-kind-icon">device<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">GPUDevice</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/webgpu.ts#L50">webgpu.ts:50</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/webgpu.ts#L50">webgpu.ts:50</a></li>
</ul>
</aside>
</section>
@@ -155,7 +155,7 @@
<div class="tsd-signature tsd-kind-icon">memory<span class="tsd-signature-symbol">:</span> <a href="memory.html" class="tsd-signature-type">Memory</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/webgpu.ts#L51">webgpu.ts:51</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/webgpu.ts#L51">webgpu.ts:51</a></li>
</ul>
</aside>
</section>
@@ -172,7 +172,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/webgpu.ts#L84">webgpu.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/webgpu.ts#L84">webgpu.ts:84</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -209,7 +209,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/webgpu.ts#L172">webgpu.ts:172</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/webgpu.ts#L172">webgpu.ts:172</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -238,7 +238,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/webgpu.ts#L67">webgpu.ts:67</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/webgpu.ts#L67">webgpu.ts:67</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/enums/argtypecode.html b/docs/reference/api/typedoc/enums/argtypecode.html
index 0db9c2f440..f931fbf22f 100644
--- a/docs/reference/api/typedoc/enums/argtypecode.html
+++ b/docs/reference/api/typedoc/enums/argtypecode.html
@@ -106,7 +106,7 @@
<div class="tsd-signature tsd-kind-icon">DLDevice<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 6</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L220">ctypes.ts:220</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L220">ctypes.ts:220</a></li>
</ul>
</aside>
</section>
@@ -116,7 +116,7 @@
<div class="tsd-signature tsd-kind-icon">Float<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 2</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L216">ctypes.ts:216</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L216">ctypes.ts:216</a></li>
</ul>
</aside>
</section>
@@ -126,7 +126,7 @@
<div class="tsd-signature tsd-kind-icon">Int<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 0</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L214">ctypes.ts:214</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L214">ctypes.ts:214</a></li>
</ul>
</aside>
</section>
@@ -136,7 +136,7 @@
<div class="tsd-signature tsd-kind-icon">Null<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 4</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L218">ctypes.ts:218</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L218">ctypes.ts:218</a></li>
</ul>
</aside>
</section>
@@ -146,7 +146,7 @@
<div class="tsd-signature tsd-kind-icon">TVMBytes<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 12</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L226">ctypes.ts:226</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L226">ctypes.ts:226</a></li>
</ul>
</aside>
</section>
@@ -156,7 +156,7 @@
<div class="tsd-signature tsd-kind-icon">TVMDLTensor<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 7</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L221">ctypes.ts:221</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L221">ctypes.ts:221</a></li>
</ul>
</aside>
</section>
@@ -166,7 +166,7 @@
<div class="tsd-signature tsd-kind-icon">TVMData<wbr>Type<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 5</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L219">ctypes.ts:219</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L219">ctypes.ts:219</a></li>
</ul>
</aside>
</section>
@@ -176,7 +176,7 @@
<div class="tsd-signature tsd-kind-icon">TVMModule<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 9</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L223">ctypes.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L223">ctypes.ts:223</a></li>
</ul>
</aside>
</section>
@@ -186,7 +186,7 @@
<div class="tsd-signature tsd-kind-icon">TVMNDArray<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 13</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L227">ctypes.ts:227</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L227">ctypes.ts:227</a></li>
</ul>
</aside>
</section>
@@ -196,7 +196,7 @@
<div class="tsd-signature tsd-kind-icon">TVMObject<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 8</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L222">ctypes.ts:222</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L222">ctypes.ts:222</a></li>
</ul>
</aside>
</section>
@@ -206,7 +206,7 @@
<div class="tsd-signature tsd-kind-icon">TVMObjectRValue<wbr>Ref<wbr>Arg<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 14</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L228">ctypes.ts:228</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L228">ctypes.ts:228</a></li>
</ul>
</aside>
</section>
@@ -216,7 +216,7 @@
<div class="tsd-signature tsd-kind-icon">TVMOpaque<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 3</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L217">ctypes.ts:217</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L217">ctypes.ts:217</a></li>
</ul>
</aside>
</section>
@@ -226,7 +226,7 @@
<div class="tsd-signature tsd-kind-icon">TVMPacked<wbr>Func<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 10</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L224">ctypes.ts:224</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L224">ctypes.ts:224</a></li>
</ul>
</aside>
</section>
@@ -236,7 +236,7 @@
<div class="tsd-signature tsd-kind-icon">TVMStr<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 11</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/c36ae1cb3/web/src/ctypes.ts#L225">ctypes.ts:225</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/2e30e772a/web/src/ctypes.ts#L225">ctypes.ts:225</a></li>
... 1561 lines suppressed ...