[tvm-site] branch asf-site updated: deploying docs (apache/tvm@52739ef8cd4499b555453d7159991316afc44301)

[tq...@apache.org]

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 8edb1192be deploying docs (apache/tvm@52739ef8cd4499b555453d7159991316afc44301)
8edb1192be is described below

commit 8edb1192be90b7024c1a07959560d3acc0c0a80d
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Wed Nov 16 19:11:25 2022 +0000

    deploying docs (apache/tvm@52739ef8cd4499b555453d7159991316afc44301)
---
 .../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_android.rst.txt  |    2 +-
 .../deploy_object_detection_pytorch.rst.txt        |    4 +-
 .../deploy_models/deploy_prequantized.rst.txt      |    6 +-
 .../deploy_prequantized_tflite.rst.txt             |    4 +-
 .../how_to/deploy_models/deploy_quantized.rst.txt  |    2 +-
 .../deploy_models/deploy_ssd_gluoncv.rst.txt       |    4 +-
 .../deploy_models/sg_execution_times.rst.txt       |   18 +-
 .../extend_tvm/bring_your_own_datatypes.rst.txt    |    2 +-
 .../how_to/extend_tvm/sg_execution_times.rst.txt   |    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                 | 2170 ++++++++++++++++----
 .../tune_network_cuda.rst.txt                      |    4 +-
 .../tune_network_x86.rst.txt                       |    4 +-
 .../tune_sparse_x86.rst.txt                        |   84 +-
 .../tune_with_autotvm/sg_execution_times.rst.txt   |    6 +-
 .../tune_with_autotvm/tune_conv2d_cuda.rst.txt     |  194 +-
 .../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  |   16 +-
 .../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 |   18 +-
 .../how_to/work_with_schedules/tensorize.rst.txt   |    2 +-
 .../tutorials/autotvm/sg_execution_times.rst.txt   |    6 +-
 .../frontend/deploy_classification.rst.txt         |    2 +-
 .../tutorials/frontend/deploy_detection.rst.txt    |    2 +-
 .../tutorials/frontend/sg_execution_times.rst.txt  |    6 +-
 .../tutorials/optimize/sg_execution_times.rst.txt  |    6 +-
 .../topic/vta/tutorials/sg_execution_times.rst.txt |    6 +-
 .../tutorial/auto_scheduler_matmul_x86.rst.txt     |    4 +-
 docs/_sources/tutorial/autotvm_matmul_x86.rst.txt  |   20 +-
 docs/_sources/tutorial/autotvm_relay_x86.rst.txt   |   56 +-
 .../tutorial/cross_compilation_and_rpc.rst.txt     |    2 +-
 docs/_sources/tutorial/intro_topi.rst.txt          |    2 +-
 docs/_sources/tutorial/sg_execution_times.rst.txt  |   20 +-
 .../tutorial/tensor_expr_get_started.rst.txt       |   46 +-
 docs/commit_hash                                   |    2 +-
 docs/how_to/compile_models/from_darknet.html       |    2 +-
 docs/how_to/compile_models/from_keras.html         |    2 +-
 docs/how_to/compile_models/from_mxnet.html         |    2 +-
 docs/how_to/compile_models/from_oneflow.html       |   17 +-
 docs/how_to/compile_models/from_pytorch.html       |   14 +-
 docs/how_to/compile_models/from_tensorflow.html    |    2 +-
 docs/how_to/compile_models/sg_execution_times.html |   22 +-
 .../deploy_models/deploy_model_on_android.html     |    2 +-
 .../deploy_object_detection_pytorch.html           |   54 +-
 docs/how_to/deploy_models/deploy_prequantized.html |    8 +-
 .../deploy_models/deploy_prequantized_tflite.html  |    4 +-
 docs/how_to/deploy_models/deploy_quantized.html    |    2 +-
 docs/how_to/deploy_models/deploy_ssd_gluoncv.html  |   35 +-
 docs/how_to/deploy_models/sg_execution_times.html  |   18 +-
 .../extend_tvm/bring_your_own_datatypes.html       |    2 +-
 docs/how_to/extend_tvm/sg_execution_times.html     |    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                    | 2169 +++++++++++++++----
 .../tune_with_autoscheduler/tune_network_cuda.html |    4 +-
 .../tune_with_autoscheduler/tune_network_x86.html  |    4 +-
 .../tune_with_autoscheduler/tune_sparse_x86.html   |   84 +-
 .../tune_with_autotvm/sg_execution_times.html      |    6 +-
 .../how_to/tune_with_autotvm/tune_conv2d_cuda.html |  194 +-
 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    |   18 +-
 docs/how_to/work_with_schedules/tensorize.html     |    2 +-
 docs/install/nnpack.html                           |   12 +-
 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   |    6 +-
 docs/tutorial/auto_scheduler_matmul_x86.html       |    4 +-
 docs/tutorial/autotvm_matmul_x86.html              |   20 +-
 docs/tutorial/autotvm_relay_x86.html               |  273 +--
 docs/tutorial/cross_compilation_and_rpc.html       |    2 +-
 docs/tutorial/intro_topi.html                      |    2 +-
 docs/tutorial/sg_execution_times.html              |   20 +-
 docs/tutorial/tensor_expr_get_started.html         |   46 +-
 126 files changed, 4655 insertions(+), 1868 deletions(-)

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 d0835e1320..978279bbae 100644
--- a/docs/_sources/how_to/compile_models/from_darknet.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_darknet.rst.txt
@@ -315,7 +315,7 @@ The process is no different from other examples.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  13.405 seconds)
+   **Total running time of the script:** ( 1 minutes  16.228 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 595ac3a7a0..790b80f493 100644
--- a/docs/_sources/how_to/compile_models/from_keras.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_keras.rst.txt
@@ -228,7 +228,7 @@ 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 964ms/step
+
    1/1 [==============================] - ETA: 0s
    1/1 [==============================] - 1s 955ms/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 f110a29d18..5eb0d30fd7 100644
--- a/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
@@ -115,7 +115,7 @@ In this section, we download a pretrained imagenet model and classify an image.
 
  .. code-block:: none
 
-    Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip6703bb06-6b2d-4042-9067-f67e09bca4e5 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+    Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip2b781813-d81c-436a-a775-b37a76c5a5e2 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 641e251484..43b90986df 100644
--- a/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
@@ -116,7 +116,7 @@ Load a pretrained OneFlow model and save model
  .. code-block:: none
 
     Downloading: "https://oneflow-public.oss-cn-beijing.aliyuncs.com/model_zoo/flowvision/classification/ResNet/resnet18.zip" to /workspace/.oneflow/flowvision_cache/resnet18.zip
-
      0%|          | 0.00/41.5M [00:00<?, ?B/s]
     15%|#5        | 6.33M/41.5M [00:00<00:00, 37.1MB/s]
     24%|##3       | 9.87M/41.5M [00:00<00:01, 29.5MB/s]
     39%|###8      | 16.0M/41.5M [00:00<00:00, 33.4MB/s]
     54%|#####3    | 22.3M/41.5M [00:00<00:00, 42.2MB/s]
     73%|#######3  | 30.3M/41.5M [00:00<00:00, 54.0MB/s]
     87%|########6 | 36.0M/41.5M [00:00<00:00, 45.8MB/s]
     98%|#########8| 40.8M/41.5M [00:01<00:00, 43.9MB/s]
    100%|##########| 41.5M/41.5M [00:01<00:00, 42.9MB/s]
+
      0%|          | 0.00/41.5M [00:00<?, ?B/s]
     15%|#5        | 6.33M/41.5M [00:00<00:01, 31.4MB/s]
     27%|##6       | 11.0M/41.5M [00:00<00:00, 38.6MB/s]
     39%|###8      | 16.0M/41.5M [00:00<00:00, 33.8MB/s]
     54%|#####3    | 22.3M/41.5M [00:00<00:00, 33.7MB/s]
     62%|######1   | 25.7M/41.5M [00:00<00:00, 31.0MB/s]
     77%|#######7  | 32.0M/41.5M [00:00<00:00, 35.2MB/s]
     85%|########5 | 35.4M/41.5M [00:01<00:00, 33.5MB/s]
     93%|#########3| 38.6M/41.5M [00:01<00:00, 32.8MB/s]
    100%|##########| 41.5M/41.5M [00:01<00:00, 34.3MB/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 8cb00b1ff8..d918e0691f 100644
--- a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
@@ -98,7 +98,7 @@ 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]
      6%|5         | 2.61M/44.7M [00:00<00:01, 22.4MB/s]
     18%|#7        | 7.99M/44.7M [00:00<00:01, 33.0MB/s]
     32%|###2      | 14.3M/44.7M [00:00<00:01, 31.7MB/s]
     39%|###8      | 17.3M/44.7M [00:00<00:00, 29.7MB/s]
     55%|#####4    | 24.5M/44.7M [00:00<00:00, 41.9MB/s]
     72%|#######1  | 32.0M/44.7M [00:00<00:00, 43.9MB/s]
     90%|########9 | 40.0M/44.7M [00:01<00:00, 42.2MB/s]
    100%|##########| 44.7M/44.7M [00:01<00:00, 42.3MB/s]
+
      0%|          | 0.00/44.7M [00:00<?, ?B/s]
     18%|#7        | 7.99M/44.7M [00:00<00:00, 68.8MB/s]
     36%|###5      | 16.0M/44.7M [00:00<00:00, 55.9MB/s]
     54%|#####3    | 24.0M/44.7M [00:00<00:00, 62.2MB/s]
     72%|#######1  | 32.0M/44.7M [00:00<00:00, 66.3MB/s]
     90%|########9 | 40.1M/44.7M [00:00<00:00, 70.1MB/s]
    100%|##########| 44.7M/44.7M [00:00<00:00, 71.9MB/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 e71f44fca2..b608418df7 100644
--- a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
@@ -416,7 +416,7 @@ Run the corresponding model on tensorflow
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  11.788 seconds)
+   **Total running time of the script:** ( 1 minutes  13.077 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 3cd540c36a..de4e8e5f7a 100644
--- a/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
@@ -5,26 +5,26 @@
 
 Computation times
 =================
-**05:52.981** total execution time for **how_to_compile_models** files:
+**05:58.074** total execution time for **how_to_compile_models** files:
 
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``)       | 01:13.405 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``)       | 01:16.228 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:11.788 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:13.077 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``)         | 00:46.658 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``)         | 00:47.284 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``)       | 00:33.757 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``)       | 00:33.815 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``)           | 00:30.473 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``)           | 00:29.966 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``)         | 00:26.763 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``)         | 00:27.561 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``)         | 00:25.790 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``)         | 00:26.150 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``)       | 00:23.270 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``)       | 00:23.095 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``)           | 00:18.636 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``)           | 00:18.467 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``)             | 00:02.439 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``)             | 00:02.433 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
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 63a1f20153..d0ab064d23 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
@@ -433,7 +433,7 @@ Execute on TVM
     Evaluate inference time cost...
     Execution time summary:
      mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)  
-      16.2805      16.3876      16.6878      15.6962       0.3367   
+      16.1529      16.1286      16.3120      16.0284       0.0906   
                
 
 
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 553acad3b6..c93c63755d 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
@@ -127,7 +127,7 @@ Load pre-trained maskrcnn from torchvision and do tracing
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=MaskRCNN_ResNet50_FPN_Weights.COCO_V1`. You can also use `weights=MaskRCNN_ResNet50_FPN_Weights.DEFAULT` to get the most up-to-date weights.
       warnings.warn(msg)
     Downloading: "https://download.pytorch.org/models/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth" to /workspace/.cache/torch/hub/checkpoints/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth
-
      0%|          | 0.00/170M [00:00<?, ?B/s]
      5%|4         | 7.99M/170M [00:00<00:02, 61.7MB/s]
      8%|8         | 14.3M/170M [00:00<00:04, 36.6MB/s]
     11%|#         | 18.2M/170M [00:00<00:04, 37.0MB/s]
     13%|#3        | 22.3M/170M [00:00<00:05, 29.5MB/s]
     15%|#4        | 25.4M/170M [00:00<00:05, 29.7MB/s]
     21%|##        | 35.3M/170M [00:00<00:02, 48.6MB/s]
     24%|##3       | 40.7M/170M [00:01<00:02, 46.7MB/s]
     28%|##8       | 48.0M/170M [00:01<00:02, 48.3MB/s]
     33%|###2      | 56.0M/170M [00:01<00:02, 52.5MB/s]
     38%|###7      | 64.0M/170M [00:01<00:01, 58.0MB/s]
     42%|####2     | 72.0M/170M [00:01<00:01, 53.5MB/s]
     47%|####7     | 80.0M/170M [00:01<00:01, 53.3MB/s]
     51%|#####     | 86.3M/170M [00:01<00:01, 56.0MB/s]
     54%|#####4    | 91.9M/170M [00:01<00:01, 56.0MB/s]
     58%|#####8    | 98.6M/170M [00:02<00:01, 59.6MB/s]
     62%|######1   | 104M/170M [00:02<00:01, 54.0MB/s] 
     68%|######7   | 115M/170M [00:02<00:00, 68.1MB/s
 ]
     72%|#######1  | 122M/170M [00:02<00:00, 66.7MB/s]
     76%|#######5  | 128M/170M [00:02<00:00, 64.8MB/s]
     80%|########  | 136M/170M [00:02<00:00, 56.4MB/s]
     83%|########3 | 142M/170M [00:02<00:00, 51.1MB/s]
     87%|########7 | 148M/170M [00:02<00:00, 54.5MB/s]
     90%|######### | 153M/170M [00:03<00:00, 42.6MB/s]
     94%|#########4| 160M/170M [00:03<00:00, 45.1MB/s]
     99%|#########8| 168M/170M [00:03<00:00, 53.5MB/s]
    100%|##########| 170M/170M [00:03<00:00, 51.6MB/s]
+
      0%|          | 0.00/170M [00:00<?, ?B/s]
      5%|4         | 7.99M/170M [00:00<00:04, 39.9MB/s]
      8%|8         | 14.3M/170M [00:00<00:03, 40.8MB/s]
     11%|#         | 18.2M/170M [00:00<00:04, 36.0MB/s]
     14%|#4        | 24.0M/170M [00:00<00:04, 36.7MB/s]
     19%|#8        | 32.0M/170M [00:00<00:03, 38.2MB/s]
     24%|##3       | 40.0M/170M [00:01<00:03, 43.5MB/s]
     28%|##8       | 48.0M/170M [00:01<00:02, 43.9MB/s]
     33%|###2      | 56.0M/170M [00:01<00:02, 48.5MB/s]
     37%|###6      | 62.3M/170M [00:01<00:02, 52.2MB/s]
     40%|###9      | 67.5M/170M [00:01<00:02, 47.1MB/s]
     42%|####2     | 72.2M/170M [00:01<00:02, 41.2MB/s]
     47%|####6     | 79.7M/170M [00:01<00:01, 49.6MB/s]
     50%|####9     | 84.8M/170M [00:02<00:01, 45.0MB/s]
     53%|#####2    | 89.5M/170M [00:02<00:02, 33.8MB/s]
     56%|#####6    | 95.5M/170M [00:02<00:01, 39.6MB/s]
     59%|#####8    | 99.9M/170M [00:02<00:02, 35.7MB/s]
     61%|######1   | 104M/170M [00:02<00:02, 31.7MB/s
 ] 
     66%|######5   | 112M/170M [00:02<00:01, 35.3MB/s]
     71%|#######   | 120M/170M [00:03<00:01, 43.0MB/s]
     75%|#######5  | 128M/170M [00:03<00:00, 44.1MB/s]
     80%|########  | 136M/170M [00:03<00:00, 46.6MB/s]
     85%|########4 | 144M/170M [00:03<00:00, 45.1MB/s]
     89%|########9 | 152M/170M [00:03<00:00, 45.9MB/s]
     93%|#########3| 158M/170M [00:03<00:00, 45.5MB/s]
     96%|#########5| 163M/170M [00:04<00:00, 41.6MB/s]
    100%|##########| 170M/170M [00:04<00:00, 42.9MB/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').
@@ -296,7 +296,7 @@ Get boxes with score larger than 0.9
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 3 minutes  16.971 seconds)
+   **Total running time of the script:** ( 3 minutes  24.737 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 a883aa64b2..f4e18c1526 100644
--- a/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
@@ -236,7 +236,7 @@ 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]
     47%|####6     | 6.30M/13.6M [00:00<00:00, 63.1MB/s]
     91%|######### | 12.3M/13.6M [00:00<00:00, 44.4MB/s]
    100%|##########| 13.6M/13.6M [00:00<00:00, 50.0MB/s]
+
      0%|          | 0.00/13.6M [00:00<?, ?B/s]
    100%|##########| 13.6M/13.6M [00:00<00:00, 159MB/s]
 
 
 
@@ -418,7 +418,7 @@ 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)  
-      90.4050      90.3277      95.0359      90.0595       0.4947   
+      90.6324      90.5561      93.1118      90.2627       0.4082   
                
 
 
@@ -467,7 +467,7 @@ TODO
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  6.767 seconds)
+   **Total running time of the script:** ( 1 minutes  8.512 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 d40eb3fa1a..0ff817b867 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
@@ -432,7 +432,7 @@ 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)  
-      120.9329     120.8553     125.1488     120.2441      0.5669   
+      121.4938     121.4546     125.6496     120.6399      0.5865   
                
 
 
@@ -469,7 +469,7 @@ 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  31.481 seconds)
+   **Total running time of the script:** ( 2 minutes  28.305 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 42da9af580..ce30a8b703 100644
--- a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
@@ -253,7 +253,7 @@ We create a Relay VM to build and execute the model.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  39.807 seconds)
+   **Total running time of the script:** ( 1 minutes  42.212 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 0d73d00e9e..e87f384719 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
@@ -166,7 +166,7 @@ Convert and compile model for CPU.
             data: None
       input_sym_arg_type = in_param.infer_type()[0]
     Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/ssd_512_resnet50_v1_voc-9c8b225a.zip...
-
      0%|          | 0/132723 [00:00<?, ?KB/s]
      6%|5         | 7359/132723 [00:00<00:01, 73578.32KB/s]
     12%|#2        | 16072/132723 [00:00<00:01, 81546.45KB/s]
     18%|#8        | 24227/132723 [00:00<00:01, 62005.86KB/s]
     25%|##4       | 32757/132723 [00:00<00:02, 43299.33KB/s]
     31%|###1      | 41408/132723 [00:00<00:01, 53118.25KB/s]
     38%|###7      | 50040/132723 [00:00<00:01, 61252.25KB/s]
     44%|####4     | 58670/132723 [00:00<00:01, 67745.91KB/s]
     51%|#####     | 67351/132723 [00:01<00:00, 72896.78KB/s]
     57%|#####7    | 75784/132723 [00:01<00:00, 76080.18KB/s]
     63%|######3   | 84228/132723 [00:01<00:00, 78457.65KB/s]
     70%|######9   | 92783/132723 [00:01<00:00, 80506.54KB/s]
     76%|#######6  | 101108/132723 [00:01<00:00, 70659.05KB/s]
     83%|########2 | 109815/132723 [00:01<00:00, 75025.98KB/s]
     89%|########8 | 118048/132723 [00:01<00:00, 77036.58KB/s]
     95%|#########5| 126742/132723 [00:01<00:00, 79831.45KB/s]
    100%|#######
 ###| 132723/132723 [00:01<00:00, 69886.02KB/s]
+
      0%|          | 0/132723 [00:00<?, ?KB/s]
      5%|4         | 6163/132723 [00:00<00:02, 61623.30KB/s]
     11%|#1        | 14890/132723 [00:00<00:01, 76706.81KB/s]
     17%|#6        | 22561/132723 [00:00<00:01, 61488.26KB/s]
     24%|##3       | 31336/132723 [00:00<00:01, 70502.66KB/s]
     29%|##9       | 39069/132723 [00:00<00:01, 72748.29KB/s]
     36%|###6      | 47847/132723 [00:00<00:01, 77552.97KB/s]
     42%|####2     | 55779/132723 [00:00<00:01, 60308.08KB/s]
     49%|####8     | 64533/132723 [00:00<00:01, 67303.49KB/s]
     54%|#####4    | 71883/132723 [00:01<00:00, 68966.09KB/s]
     61%|######    | 80613/132723 [00:01<00:00, 74045.10KB/s]
     67%|######7   | 89317/132723 [00:01<00:00, 77730.07KB/s]
     74%|#######3  | 97956/132723 [00:01<00:00, 80224.79KB/s]
     80%|########  | 106638/132723 [00:01<00:00, 82146.77KB/s]
     87%|########6 | 115007/132723 [00:01<00:00, 66534.79KB/s]
     93%|#########3| 123756/132723 [00:01<00:00, 71833.44KB/s]
     99%|########
 #9| 131450/132723 [00:01<00:00, 73184.77KB/s]
    100%|##########| 132723/132723 [00:01<00:00, 71855.54KB/s]
 
 
 
@@ -242,7 +242,7 @@ Display result
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 3 minutes  2.439 seconds)
+   **Total running time of the script:** ( 3 minutes  7.321 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 457a57626d..be5d619c0c 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
@@ -5,24 +5,24 @@
 
 Computation times
 =================
-**13:04.439** total execution time for **how_to_deploy_models** files:
+**13:21.255** total execution time for **how_to_deploy_models** files:
 
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:16.971 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:24.737 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``)                           | 03:02.439 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``)                           | 03:07.321 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``)           | 02:31.481 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``)           | 02:28.305 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``)                               | 01:39.807 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``)                               | 01:42.212 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``)                         | 01:06.767 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``)                         | 01:08.512 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``)                 | 00:36.121 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``)                 | 00:37.861 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``)                       | 00:25.702 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``)                       | 00:26.511 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``)                       | 00:25.146 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``)                       | 00:25.790 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_deploy_models_deploy_sparse.py` (``deploy_sparse.py``)                                     | 00:00.007 | 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 4083b3b8a3..e3c6121d81 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
@@ -472,7 +472,7 @@ 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.zipb5ec6e0e-1a2e-4ead-b5c9-a1715a34eb53 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+    Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip85164043-6ab1-4ce7-9f25-a44176bc77c5 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 170b276518..2cfb8109ed 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
 =================
-**00:48.729** total execution time for **how_to_extend_tvm** files:
+**00:49.063** 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``) | 00:45.240 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:45.485 | 0.0 MB |
 +-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``)           | 00:02.441 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``)           | 00:02.505 | 0.0 MB |
 +-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``)                     | 00:01.040 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``)                     | 00:01.065 | 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 82afb7faa1..64f94f31d2 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
@@ -216,10 +216,10 @@ profile the execution time of each passes.
  .. code-block:: none
 
     Printing results of timing profile...
-    InferType: 7091us [7091us] (45.25%; 45.25%)
-    FoldScaleAxis: 8579us [7us] (54.75%; 54.75%)
-            FoldConstant: 8572us [1775us] (54.70%; 99.92%)
-                    InferType: 6796us [6796us] (43.37%; 79.29%)
+    InferType: 7311us [7311us] (46.27%; 46.27%)
+    FoldScaleAxis: 8490us [8us] (53.73%; 53.73%)
+            FoldConstant: 8482us [1698us] (53.68%; 99.90%)
+                    InferType: 6783us [6783us] (42.93%; 79.98%)
 
 
 
@@ -258,10 +258,10 @@ Refer to following sections and :py:func:`tvm.instrument.pass_instrument` for th
  .. code-block:: none
 
     Printing results of timing profile...
-    InferType: 6623us [6623us] (44.82%; 44.82%)
-    FoldScaleAxis: 8153us [5us] (55.18%; 55.18%)
-            FoldConstant: 8148us [1692us] (55.14%; 99.94%)
-                    InferType: 6456us [6456us] (43.69%; 79.24%)
+    InferType: 6902us [6902us] (44.90%; 44.90%)
+    FoldScaleAxis: 8470us [6us] (55.10%; 55.10%)
+            FoldConstant: 8464us [1732us] (55.06%; 99.93%)
+                    InferType: 6732us [6732us] (43.79%; 79.53%)
 
 
 
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 8edfefa1cd..4923070246 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
@@ -340,7 +340,7 @@ latency of convolution.
 
  .. code-block:: none
 
-    Convolution: 34.213375 ms
+    Convolution: 54.245216 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 d266275182..8e9189bc7d 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
@@ -657,7 +657,7 @@ be able to run on our build server
 
  .. code-block:: none
 
-    conv2d with tensor core: 13.353340 ms
+    conv2d with tensor core: 13.002118 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 9271533431..e0f9d1693a 100644
--- a/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
@@ -143,8 +143,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
 
  .. code-block:: none
 
-    Numpy running time: 0.019583
-    Baseline: 3.429270
+    Numpy running time: 0.019491
+    Baseline: 3.290225
 
 
 
@@ -238,7 +238,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
 
  .. code-block:: none
 
-    Opt1: 0.308873
+    Opt1: 0.328238
 
 
 
@@ -340,7 +340,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
 
  .. code-block:: none
 
-    Opt2: 0.331130
+    Opt2: 0.353639
 
 
 
@@ -435,7 +435,7 @@ the access pattern for A matrix is more cache friendly.
 
  .. code-block:: none
 
-    Opt3: 0.116499
+    Opt3: 0.122369
 
 
 
@@ -559,7 +559,7 @@ flattening.
 
  .. code-block:: none
 
-    Opt4: 0.109366
+    Opt4: 0.109637
 
 
 
@@ -680,7 +680,7 @@ write to C when all the block results are ready.
 
  .. code-block:: none
 
-    Opt5: 0.111082
+    Opt5: 0.111313
 
 
 
@@ -804,7 +804,7 @@ Furthermore, we can also utilize multi-core processors to do the thread-level pa
 
  .. code-block:: none
 
-    Opt6: 0.147840
+    Opt6: 0.147555
 
 
 
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 cf302c36db..532377169b 100644
--- a/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
@@ -5,12 +5,12 @@
 
 Computation times
 =================
-**00:35.433** total execution time for **how_to_optimize_operators** files:
+**00:35.420** total execution time for **how_to_optimize_operators** files:
 
 +-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``)                       | 00:32.596 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``)                       | 00:32.731 | 0.0 MB |
 +-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.698 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.512 | 0.0 MB |
 +-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``)             | 00:01.138 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``)             | 00:01.176 | 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 ba6acac68a..030d97cdb2 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
 =================
-**09:22.962** total execution time for **how_to_tune_with_autoscheduler** files:
+**09:16.506** 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``) | 05:52.239 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 05:43.754 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``)             | 01:34.879 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``)             | 01:34.813 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``)           | 01:04.373 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``)           | 01:04.752 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``)               | 00:27.840 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``)               | 00:28.883 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``)             | 00:12.218 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``)             | 00:12.495 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``)           | 00:11.414 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``)           | 00:11.809 | 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 8579666368..99318a7090 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
@@ -239,191 +239,952 @@ cooperative fetching, unrolling and operator fusion.
                  bias: Buffer(bias_2: Pointer(float32), float32, [1, 512, 1, 1], []),
                  compute: Buffer(compute_2: Pointer(float32), float32, [1, 512, 7, 7], [])}
       buffer_map = {data_1: data, kernel_1: kernel, bias_1: bias, compute_1: compute} {
-      attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 128;
-      allocate(conv2d_nchw: Pointer(local float32), float32, [4]), storage_scope = local;
-      allocate(pad_temp.shared: Pointer(shared float32), float32, [324]), storage_scope = shared;
-      allocate(kernel.shared: Pointer(shared float32), float32, [144]), storage_scope = shared;
+      attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 32;
+      allocate(conv2d_nchw: Pointer(local float32), float32, [16]), storage_scope = local;
+      allocate(pad_temp.shared: Pointer(shared float32), float32, [784]), storage_scope = shared;
+      allocate(kernel.shared: Pointer(shared float32), float32, [256]), storage_scope = shared;
       attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
-        conv2d_nchw_1: Buffer(conv2d_nchw, float32, [4], [], scope="local", align=16)[0] = 0f32
-        conv2d_nchw_1[1] = 0f32
+        conv2d_nchw_1: Buffer(conv2d_nchw, float32, [4], [], scope="local", align=8)[0] = 0f32
         conv2d_nchw_1[2] = 0f32
+        conv2d_nchw_1[4] = 0f32
+        conv2d_nchw_1[6] = 0f32
+        conv2d_nchw_1[8] = 0f32
+        conv2d_nchw_1[10] = 0f32
+        conv2d_nchw_1[12] = 0f32
+        conv2d_nchw_1[14] = 0f32
+        conv2d_nchw_1[1] = 0f32
         conv2d_nchw_1[3] = 0f32
-        for (rc.outer.outer: int32, 0, 128) {
-          let cse_var_2: int32 = (rc.outer.outer*196)
-          let cse_var_1: int32 = (rc.outer.outer*36)
-           {
+        conv2d_nchw_1[5] = 0f32
+        conv2d_nchw_1[7] = 0f32
+        conv2d_nchw_1[9] = 0f32
+        conv2d_nchw_1[11] = 0f32
+        conv2d_nchw_1[13] = 0f32
+        conv2d_nchw_1[15] = 0f32
+        for (rc.outer.outer: int32, 0, 32) {
+          for (ry.outer.outer: int32, 0, 3) {
             attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            pad_temp.shared_1: Buffer(pad_temp.shared, float32, [324], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else((((9 <= threadIdx.x_1) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data_3: Buffer(data_2, float32, [25088], [])[(((cse_var_2 + (floordiv(threadIdx.x_1, 9)*7)) + floormod(threadIdx.x_1, 9)) - 8)], 0f32, dtype=float32)
+            pad_temp.shared_1: Buffer(pad_temp.shared, float32, [784], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3: Buffer(data_2, float32, [25088], [])[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) - 8)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 41)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 90)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 139)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 188)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 237)], 0f32, dtype=float32)
             attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 49), 81)) && (floormod((threadIdx.x_1 + 49), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 4), 9))) && (floormod((threadIdx.x_1 + 4), 9) < 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 49), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 49), 81), 9)*7)) + floormod((threadIdx.x_1 + 4), 9)) - 8)], 0f32, dtype=float32)
+            pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 286)], 0f32, dtype=float32)
             attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else(((1 <= floormod((threadIdx.x_1 + 8), 9)) && (floormod((threadIdx.x_1 + 8), 9) < 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 98), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 17), 81), 9)*7)) + floormod((threadIdx.x_1 + 8), 9)) - 8)], 0f32, dtype=float32)
+            pad_temp.shared_1[(threadIdx.x_1 + 343)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 335)], 0f32, dtype=float32)
             attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 66), 81)) && (floormod((threadIdx.x_1 + 66), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 3), 9))) && (floormod((threadIdx.x_1 + 3), 9) < 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 147), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 66), 81), 9)*7)) + floormod((threadIdx.x_1 + 3), 9)) - 8)], 0f32, dtype=float32)
+            pad_temp.shared_1[(threadIdx.x_1 + 392)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 384)], 0f32, dtype=float32)
             attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 34), 81)) && (floormod((threadIdx.x_1 + 34), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 7), 9))) && (floormod((threadIdx.x_1 + 7), 9) < 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 196), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 34), 81), 9)*7)) + floormod((threadIdx.x_1 + 7), 9)) - 8)], 0f32, dtype=float32)
+            pad_temp.shared_1[(threadIdx.x_1 + 441)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 433)], 0f32, dtype=float32)
             attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else((((9 <= floormod((threadIdx.x_1 + 2), 81)) && (1 <= floormod((threadIdx.x_1 + 2), 9))) && (floormod((threadIdx.x_1 + 2), 9) < 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 245), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 2), 81), 9)*7)) + floormod((threadIdx.x_1 + 2), 9)) - 8)], 0f32, dtype=float32)
+            pad_temp.shared_1[(threadIdx.x_1 + 490)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 482)], 0f32, dtype=float32)
             attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            if @tir.likely((threadIdx.x_1 < 30), dtype=bool) {
-              pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else((((threadIdx.x_1 < 21) && (1 <= floormod((threadIdx.x_1 + 6), 9))) && (floormod((threadIdx.x_1 + 6), 9) < 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 294), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 51), 81), 9)*7)) + floormod((threadIdx.x_1 + 6), 9)) - 8)], 0f32, dtype=float32)
+            pad_temp.shared_1[(threadIdx.x_1 + 539)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 531)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 588)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 580)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 637)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 629)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 686)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 678)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 735)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (1 <= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 727)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+              kernel.shared_1: Buffer(kernel.shared, float32, [256], [], scope="shared")[(threadIdx.x_2*4)] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[(((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3))]
+              kernel.shared_1[((threadIdx.x_2*4) + 1)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 9)]
+              kernel.shared_1[((threadIdx.x_2*4) + 2)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 18)]
+              kernel.shared_1[((threadIdx.x_2*4) + 3)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 27)]
             }
-            attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            kernel.shared_1: Buffer(kernel.shared, float32, [144], [], scope="shared")[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[((((blockIdx.x*18432) + (floordiv(threadIdx.x_2, 36)*4608)) + cse_var_1) + floormod(threadIdx.x_2, 36))]
             attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            kernel.shared_1[(threadIdx.x_2 + 49)] = kernel_3[((((((blockIdx.x*18432) + (floordiv((threadIdx.x_2 + 49), 36)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 13), 36), 9)*9)) + (floordiv(floormod((threadIdx.x_2 + 4), 9), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            if @tir.likely((threadIdx.x_2 < 15), dtype=bool) {
+              let cse_var_1: int32 = (ry.outer.outer*3)
+               {
+                kernel.shared_1[((threadIdx.x_2*4) + 196)] = kernel_3[(((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 4), 16)*9)) + cse_var_1)]
+                kernel.shared_1[((threadIdx.x_2*4) + 197)] = kernel_3[(((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 5), 16)*9)) + cse_var_1)]
+                kernel.shared_1[((threadIdx.x_2*4) + 198)] = kernel_3[(((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 6), 16)*9)) + cse_var_1)]
+                kernel.shared_1[((threadIdx.x_2*4) + 199)] = kernel_3[(((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 7), 16)*9)) + cse_var_1)]
+              }
+            }
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[0]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[32]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[64]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[96]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[128]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[160]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[192]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[224]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[16]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[48]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[80]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[112]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[144]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[176]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[208]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[240]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[1]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[33]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[65]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[97]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[129]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[161]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[193]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[225]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[17]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[49]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[81]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[113]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[145]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[177]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[209]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[241]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[2]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[34]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[66]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[98]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[130]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[162]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[194]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[226]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[18]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[50]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[82]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[114]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[146]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[178]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[210]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[242]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[3]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[35]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[67]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[99]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[131]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[163]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[195]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[227]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[19]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[51]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[83]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[115]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[147]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[179]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[211]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[243]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[4]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[36]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[68]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[100]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[132]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[164]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[196]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[228]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[20]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[52]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[84]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[116]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[148]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[180]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[212]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[244]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[5]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[37]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[69]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[101]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[133]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[165]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[197]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[229]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[21]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[53]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[85]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[117]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[149]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[181]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[213]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[245]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[6]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[38]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[70]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[102]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[134]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[166]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[198]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[230]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[22]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[54]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[86]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[118]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[150]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[182]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[214]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[246]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[7]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[39]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[71]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[103]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[135]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[167]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[199]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[231]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[23]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[55]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[87]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[119]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[151]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[183]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[215]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[247]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[8]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[40]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[72]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[104]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[136]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[168]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[200]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[232]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[24]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[56]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[88]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[120]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[152]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[184]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[216]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[248]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[9]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[41]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[73]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[105]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[137]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[169]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[201]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[233]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[25]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[57]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[89]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[121]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[153]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[185]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[217]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[249]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[10]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[42]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[74]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[106]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[138]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[170]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[202]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[234]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[26]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[58]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[90]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[122]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[154]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[186]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[218]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[250]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[11]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[43]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[75]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[107]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[139]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[171]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[203]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[235]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[27]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[59]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[91]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[123]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[155]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[187]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[219]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[251]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[12]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[44]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[76]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[108]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[140]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[172]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[204]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[236]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[28]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[60]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[92]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[124]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[156]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[188]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[220]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[252]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[13]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[45]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[77]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[109]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[141]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[173]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[205]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[237]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[29]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[61]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[93]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[125]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[157]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[189]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[221]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[253]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[14]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[46]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[78]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[110]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[142]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[174]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[206]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[238]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[30]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[62]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[94]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[126]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[158]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[190]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[222]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[254]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[15]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[47]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[79]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[111]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[143]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[175]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[207]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[239]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[31]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[63]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[95]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[127]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[159]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[191]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[223]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[255]))
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[threadIdx.x_1] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) - 7)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 42)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 91)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 140)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 189)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 238)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 287)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 343)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 336)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 392)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 385)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 441)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 434)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 490)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 483)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 539)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 532)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 588)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 581)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 637)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 630)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 686)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 679)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 735)] = @tir.if_then_else(((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 728)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+              kernel.shared_1[(threadIdx.x_2*4)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 1)]
+              kernel.shared_1[((threadIdx.x_2*4) + 1)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 10)]
+              kernel.shared_1[((threadIdx.x_2*4) + 2)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 19)]
+              kernel.shared_1[((threadIdx.x_2*4) + 3)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 28)]
+            }
             attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
-            if @tir.likely((threadIdx.x_2 < 46), dtype=bool) {
-              kernel.shared_1[(threadIdx.x_2 + 98)] = kernel_3[((((((blockIdx.x*18432) + (floordiv((threadIdx.x_2 + 98), 36)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 26), 36), 9)*9)) + (floordiv(floormod((threadIdx.x_2 + 8), 9), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            if @tir.likely((threadIdx.x_2 < 15), dtype=bool) {
+              let cse_var_2: int32 = (ry.outer.outer*3)
+               {
+                kernel.shared_1[((threadIdx.x_2*4) + 196)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 4), 16)*9)) + cse_var_2) + 1)]
+                kernel.shared_1[((threadIdx.x_2*4) + 197)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 5), 16)*9)) + cse_var_2) + 1)]
+                kernel.shared_1[((threadIdx.x_2*4) + 198)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 6), 16)*9)) + cse_var_2) + 1)]
+                kernel.shared_1[((threadIdx.x_2*4) + 199)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 7), 16)*9)) + cse_var_2) + 1)]
+              }
             }
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7))]*kernel.shared_1[0]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[9]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7))]*kernel.shared_1[36]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[45]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7))]*kernel.shared_1[72]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[81]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7))]*kernel.shared_1[108]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[117]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[1]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 82)]*kernel.shared_1[10]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[37]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 82)]*kernel.shared_1[46]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[73]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 82)]*kernel.shared_1[82]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[109]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 82)]*kernel.shared_1[118]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[2]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 83)]*kernel.shared_1[11]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[38]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 83)]*kernel.shared_1[47]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[74]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 83)]*kernel.shared_1[83]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[110]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 83)]*kernel.shared_1[119]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[3]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[12]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[39]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[48]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[75]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[84]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[111]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[120]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 10)]*kernel.shared_1[4]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 91)]*kernel.shared_1[13]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 10)]*kernel.shared_1[40]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 91)]*kernel.shared_1[49]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 10)]*kernel.shared_1[76]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 91)]*kernel.shared_1[85]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 10)]*kernel.shared_1[112]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 91)]*kernel.shared_1[121]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 11)]*kernel.shared_1[5]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 92)]*kernel.shared_1[14]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 11)]*kernel.shared_1[41]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 92)]*kernel.shared_1[50]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 11)]*kernel.shared_1[77]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 92)]*kernel.shared_1[86]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 11)]*kernel.shared_1[113]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 92)]*kernel.shared_1[122]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[6]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[15]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[42]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[51]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[78]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[87]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[114]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[123]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 19)]*kernel.shared_1[7]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 100)]*kernel.shared_1[16]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 19)]*kernel.shared_1[43]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 100)]*kernel.shared_1[52]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 19)]*kernel.shared_1[79]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 100)]*kernel.shared_1[88]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 19)]*kernel.shared_1[115]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 100)]*kernel.shared_1[124]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 20)]*kernel.shared_1[8]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 101)]*kernel.shared_1[17]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 20)]*kernel.shared_1[44]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 101)]*kernel.shared_1[53]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 20)]*kernel.shared_1[80]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 101)]*kernel.shared_1[89]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 20)]*kernel.shared_1[116]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 101)]*kernel.shared_1[125]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[18]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[27]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[54]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[63]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[90]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[99]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[126]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[135]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 163)]*kernel.shared_1[19]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 244)]*kernel.shared_1[28]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 163)]*kernel.shared_1[55]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 244)]*kernel.shared_1[64]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 163)]*kernel.shared_1[91]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 244)]*kernel.shared_1[100]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 163)]*kernel.shared_1[127]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 244)]*kernel.shared_1[136]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 164)]*kernel.shared_1[20]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 245)]*kernel.shared_1[29]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 164)]*kernel.shared_1[56]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 245)]*kernel.shared_1[65]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 164)]*kernel.shared_1[92]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 245)]*kernel.shared_1[101]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 164)]*kernel.shared_1[128]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 245)]*kernel.shared_1[137]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[21]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[30]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[57]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[66]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[93]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[102]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[129]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[138]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 172)]*kernel.shared_1[22]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 253)]*kernel.shared_1[31]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 172)]*kernel.shared_1[58]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 253)]*kernel.shared_1[67]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 172)]*kernel.shared_1[94]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 253)]*kernel.shared_1[103]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 172)]*kernel.shared_1[130]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 253)]*kernel.shared_1[139]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 173)]*kernel.shared_1[23]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 254)]*kernel.shared_1[32]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 173)]*kernel.shared_1[59]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 254)]*kernel.shared_1[68]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 173)]*kernel.shared_1[95]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 254)]*kernel.shared_1[104]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 173)]*kernel.shared_1[131]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 254)]*kernel.shared_1[140]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[24]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[33]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[60]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[69]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[96]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[105]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[132]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[141]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 181)]*kernel.shared_1[25]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 262)]*kernel.shared_1[34]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 181)]*kernel.shared_1[61]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 262)]*kernel.shared_1[70]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 181)]*kernel.shared_1[97]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 262)]*kernel.shared_1[106]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 181)]*kernel.shared_1[133]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 262)]*kernel.shared_1[142]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 182)]*kernel.shared_1[26]))
-            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 263)]*kernel.shared_1[35]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 182)]*kernel.shared_1[62]))
-            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 263)]*kernel.shared_1[71]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 182)]*kernel.shared_1[98]))
-            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 263)]*kernel.shared_1[107]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 182)]*kernel.shared_1[134]))
-            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 263)]*kernel.shared_1[143]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[0]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[32]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[64]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[96]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[128]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[160]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[192]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[224]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[16]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[48]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[80]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[112]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[144]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[176]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[208]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[240]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[1]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[33]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[65]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[97]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[129]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[161]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[193]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[225]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[17]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[49]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[81]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[113]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[145]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[177]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[209]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[241]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[2]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[34]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[66]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[98]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[130]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[162]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[194]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[226]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[18]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[50]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[82]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[114]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[146]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[178]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[210]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[242]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[3]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[35]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[67]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[99]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[131]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[163]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[195]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[227]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[19]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[51]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[83]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[115]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[147]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[179]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[211]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[243]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[4]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[36]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[68]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[100]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[132]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[164]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[196]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[228]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[20]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[52]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[84]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[116]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[148]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[180]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[212]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[244]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[5]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[37]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[69]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[101]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[133]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[165]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[197]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[229]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[21]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[53]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[85]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[117]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[149]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[181]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[213]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[245]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[6]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[38]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[70]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[102]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[134]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[166]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[198]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[230]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[22]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[54]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[86]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[118]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[150]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[182]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[214]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[246]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[7]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[39]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[71]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[103]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[135]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[167]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[199]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[231]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[23]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[55]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[87]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[119]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[151]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[183]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[215]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[247]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[8]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[40]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[72]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[104]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[136]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[168]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[200]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[232]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[24]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[56]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[88]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[120]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[152]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[184]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[216]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[248]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[9]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[41]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[73]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[105]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[137]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[169]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[201]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[233]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[25]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[57]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[89]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[121]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[153]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[185]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[217]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[249]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[10]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[42]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[74]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[106]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[138]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[170]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[202]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[234]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[26]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[58]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[90]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[122]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[154]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[186]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[218]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[250]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[11]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[43]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[75]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[107]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[139]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[171]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[203]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[235]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[27]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[59]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[91]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[123]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[155]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[187]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[219]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[251]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[12]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[44]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[76]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[108]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[140]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[172]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[204]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[236]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[28]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[60]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[92]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[124]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[156]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[188]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[220]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[252]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[13]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[45]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[77]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[109]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[141]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[173]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[205]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[237]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[29]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[61]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[93]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[125]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[157]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[189]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[221]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[253]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[14]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[46]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[78]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[110]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[142]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[174]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[206]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[238]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[30]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[62]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[94]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[126]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[158]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[190]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[222]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[254]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[15]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[47]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[79]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[111]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[143]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[175]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[207]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[239]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[31]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[63]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[95]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[127]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[159]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[191]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[223]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[255]))
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[threadIdx.x_1] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) - 6)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 43)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 92)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 141)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 190)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 239)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 288)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 343)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 337)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 392)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 386)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 441)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 435)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 490)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 484)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 539)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 533)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 588)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 582)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 637)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 631)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 686)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 680)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            pad_temp.shared_1[(threadIdx.x_1 + 735)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) && ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) < 8)) && (floormod(threadIdx.x_1, 7) < 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 729)], 0f32, dtype=float32)
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+              kernel.shared_1[(threadIdx.x_2*4)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 2)]
+              kernel.shared_1[((threadIdx.x_2*4) + 1)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 11)]
+              kernel.shared_1[((threadIdx.x_2*4) + 2)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 20)]
+              kernel.shared_1[((threadIdx.x_2*4) + 3)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 29)]
+            }
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+            if @tir.likely((threadIdx.x_2 < 15), dtype=bool) {
+              let cse_var_3: int32 = (ry.outer.outer*3)
+               {
+                kernel.shared_1[((threadIdx.x_2*4) + 196)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 4), 16)*9)) + cse_var_3) + 2)]
+                kernel.shared_1[((threadIdx.x_2*4) + 197)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 5), 16)*9)) + cse_var_3) + 2)]
+                kernel.shared_1[((threadIdx.x_2*4) + 198)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 6), 16)*9)) + cse_var_3) + 2)]
+                kernel.shared_1[((threadIdx.x_2*4) + 199)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 7), 16)*9)) + cse_var_3) + 2)]
+              }
+            }
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[0]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[32]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[64]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[96]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[128]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[160]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[192]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[224]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[16]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[48]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[80]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[112]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[144]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[176]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[208]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[240]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[1]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[33]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[65]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[97]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[129]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[161]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[193]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[225]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[17]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[49]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[81]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[113]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[145]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[177]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[209]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[241]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[2]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[34]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[66]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[98]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[130]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[162]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[194]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[226]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[18]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[50]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[82]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[114]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[146]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[178]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[210]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[242]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[3]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[35]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[67]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[99]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[131]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[163]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[195]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[227]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[19]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[51]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[83]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[115]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[147]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[179]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[211]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[243]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[4]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[36]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[68]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[100]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[132]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[164]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[196]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[228]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[20]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[52]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[84]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[116]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[148]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[180]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[212]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[244]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[5]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[37]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[69]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[101]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[133]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[165]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[197]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[229]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[21]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[53]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[85]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[117]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[149]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[181]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[213]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[245]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[6]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[38]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[70]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[102]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[134]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[166]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[198]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[230]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[22]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[54]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[86]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[118]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[150]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[182]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[214]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[246]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[7]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[39]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[71]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[103]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[135]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[167]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[199]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[231]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[23]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[55]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[87]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[119]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[151]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[183]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[215]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[247]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[8]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[40]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[72]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[104]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[136]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[168]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[200]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[232]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[24]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[56]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[88]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[120]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[152]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[184]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[216]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[248]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[9]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[41]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[73]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[105]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[137]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[169]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[201]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[233]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[25]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[57]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[89]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[121]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[153]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[185]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[217]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[249]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[10]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[42]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[74]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[106]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[138]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[170]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[202]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[234]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[26]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[58]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[90]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[122]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[154]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[186]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[218]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[250]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[11]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[43]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[75]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[107]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[139]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[171]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[203]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[235]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[27]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[59]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[91]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[123]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[155]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[187]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[219]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[251]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[12]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[44]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[76]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[108]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[140]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[172]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[204]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[236]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[28]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[60]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[92]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[124]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[156]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[188]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[220]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[252]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[13]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[45]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[77]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[109]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[141]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[173]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[205]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[237]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[29]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[61]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[93]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[125]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[157]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[189]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[221]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[253]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[14]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[46]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[78]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[110]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[142]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[174]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[206]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[238]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[30]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[62]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[94]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[126]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[158]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[190]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[222]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[254]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[15]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[47]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[79]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[111]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[143]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[175]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[207]))
+            conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[239]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[31]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[63]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[95]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[127]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[159]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[191]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[223]))
+            conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[255]))
           }
         }
-        for (i1.inner: int32, 0, 4) {
-          compute_3: Buffer(compute_2, float32, [25088], [])[(((blockIdx.x*196) + (i1.inner*49)) + threadIdx.x)] = max((conv2d_nchw_1[i1.inner] + bias_3: Buffer(bias_2, float32, [512], [])[((blockIdx.x*4) + i1.inner)]), 0f32)
+        for (i1.inner: int32, 0, 2) {
+          compute_3: Buffer(compute_2, float32, [25088], [])[(((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x)] = max((conv2d_nchw_1[i1.inner] + bias_3: Buffer(bias_2, float32, [512], [])[((blockIdx.x*16) + i1.inner)]), 0f32)
+          compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 98)] = max((conv2d_nchw_1[(i1.inner + 2)] + bias_3[(((blockIdx.x*16) + i1.inner) + 2)]), 0f32)
+          compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 196)] = max((conv2d_nchw_1[(i1.inner + 4)] + bias_3[(((blockIdx.x*16) + i1.inner) + 4)]), 0f32)
+          compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 294)] = max((conv2d_nchw_1[(i1.inner + 6)] + bias_3[(((blockIdx.x*16) + i1.inner) + 6)]), 0f32)
+          compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 392)] = max((conv2d_nchw_1[(i1.inner + 8)] + bias_3[(((blockIdx.x*16) + i1.inner) + 8)]), 0f32)
+          compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 490)] = max((conv2d_nchw_1[(i1.inner + 10)] + bias_3[(((blockIdx.x*16) + i1.inner) + 10)]), 0f32)
+          compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 588)] = max((conv2d_nchw_1[(i1.inner + 12)] + bias_3[(((blockIdx.x*16) + i1.inner) + 12)]), 0f32)
+          compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 686)] = max((conv2d_nchw_1[(i1.inner + 14)] + bias_3[(((blockIdx.x*16) + i1.inner) + 14)]), 0f32)
         }
       }
     }
@@ -478,7 +1239,7 @@ We build the binary and check its correctness and performance.
 
  .. code-block:: none
 
-    Execution time of this operator: 0.259 ms
+    Execution time of this operator: 0.355 ms
 
 
 
@@ -526,10 +1287,10 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
     conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_i, factor=1)
     conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
     conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
-    conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=1)
-    conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=4)
+    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=1)
-    conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
+    conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=8)
     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)
@@ -538,19 +1299,19 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
     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=2)
+    conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=8)
     conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=2)
     conv2d_nchw_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_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
     conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=1)
-    conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=3)
+    conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=1)
     s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2 [...]
     compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
     compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
     compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
-    compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=4)
+    compute_i1_o_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=1)
-    compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+    compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=8)
     compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
     compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
     compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
@@ -573,7 +1334,7 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
     compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused = s[compute].fuse(compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i)
     s[compute].bind(compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused, te.thread_axis("threadIdx.x"))
     kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[kernel_shared].fuse(kernel_shared_ax0, kernel_shared_ax1, kernel_shared_ax2, kernel_shared_ax3)
-    kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
+    kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=4)
     s[kernel_shared].vectorize(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
     kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=49)
     s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
@@ -582,7 +1343,7 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
     s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
     pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=49)
     s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
-    s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
+    s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 1024)
     s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
 
     CUDA source code:
@@ -601,177 +1362,890 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
       #define uint64_t unsigned long long
     #endif
     extern "C" __global__ void __launch_bounds__(49) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
-      float conv2d_nchw[4];
-      __shared__ float pad_temp_shared[324];
-      __shared__ float kernel_shared[144];
+      float conv2d_nchw[16];
+      __shared__ float pad_temp_shared[784];
+      __shared__ float kernel_shared[256];
       conv2d_nchw[0] = 0.000000e+00f;
-      conv2d_nchw[1] = 0.000000e+00f;
       conv2d_nchw[2] = 0.000000e+00f;
+      conv2d_nchw[4] = 0.000000e+00f;
+      conv2d_nchw[6] = 0.000000e+00f;
+      conv2d_nchw[8] = 0.000000e+00f;
+      conv2d_nchw[10] = 0.000000e+00f;
+      conv2d_nchw[12] = 0.000000e+00f;
+      conv2d_nchw[14] = 0.000000e+00f;
+      conv2d_nchw[1] = 0.000000e+00f;
       conv2d_nchw[3] = 0.000000e+00f;
-      for (int rc_outer_outer = 0; rc_outer_outer < 128; ++rc_outer_outer) {
-        __syncthreads();
-        pad_temp_shared[((int)threadIdx.x)] = ((((9 <= ((int)threadIdx.x)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((rc_outer_outer * 196) + ((((int)threadIdx.x) / 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
-        pad_temp_shared[(((int)threadIdx.x) + 49)] = (((((9 <= ((((int)threadIdx.x) + 49) % 81)) && (((((int)threadIdx.x) + 49) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 49) / 81) * 49)) + ((((((int)threadIdx.x) + 49) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
-        pad_temp_shared[(((int)threadIdx.x) + 98)] = (((1 <= ((((int)threadIdx.x) + 8) % 9)) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 98) / 81) * 49)) + (((((int)threadIdx.x) + 17) / 9) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
-        pad_temp_shared[(((int)threadIdx.x) + 147)] = (((((9 <= ((((int)threadIdx.x) + 66) % 81)) && (((((int)threadIdx.x) + 66) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 147) / 81) * 49)) + ((((((int)threadIdx.x) + 66) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
-        pad_temp_shared[(((int)threadIdx.x) + 196)] = (((((9 <= ((((int)threadIdx.x) + 34) % 81)) && (((((int)threadIdx.x) + 34) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 196) / 81) * 49)) + ((((((int)threadIdx.x) + 34) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] : 0.000000e+00f);
-        pad_temp_shared[(((int)threadIdx.x) + 245)] = ((((7 <= ((int)threadIdx.x)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 245) / 81) * 49)) + (((((int)threadIdx.x) + 2) / 9) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
-        if (((int)threadIdx.x) < 30) {
-          pad_temp_shared[(((int)threadIdx.x) + 294)] = ((((((int)threadIdx.x) < 21) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 294) / 81) * 49)) + (((((int)threadIdx.x) + 51) / 9) * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0.000000e+00f);
-        }
-        kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) * 18432) + ((((int)threadIdx.x) / 36) * 4608)) + (rc_outer_outer * 36)) + (((int)threadIdx.x) % 36))];
-        kernel_shared[(((int)threadIdx.x) + 49)] = kernel[((((((((int)blockIdx.x) * 18432) + (((((int)threadIdx.x) + 49) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((((int)threadIdx.x) + 13) % 36) / 9) * 9)) + ((((((int)threadIdx.x) + 4) % 9) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-        if (((int)threadIdx.x) < 46) {
-          kernel_shared[(((int)threadIdx.x) + 98)] = kernel[((((((((int)blockIdx.x) * 18432) + (((((int)threadIdx.x) + 98) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((((int)threadIdx.x) + 26) % 36) / 9) * 9)) + ((((((int)threadIdx.x) + 8) % 9) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+      conv2d_nchw[5] = 0.000000e+00f;
+      conv2d_nchw[7] = 0.000000e+00f;
+      conv2d_nchw[9] = 0.000000e+00f;
+      conv2d_nchw[11] = 0.000000e+00f;
+      conv2d_nchw[13] = 0.000000e+00f;
+      conv2d_nchw[15] = 0.000000e+00f;
+      for (int rc_outer_outer = 0; rc_outer_outer < 32; ++rc_outer_outer) {
+        for (int ry_outer_outer = 0; ry_outer_outer < 3; ++ry_outer_outer) {
+          __syncthreads();
+          pad_temp_shared[((int)threadIdx.x)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) - 8)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 49)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 41)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 98)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 90)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 147)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 139)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 196)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 188)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 245)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 237)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 294)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 286)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 343)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 335)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 392)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 384)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 441)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 433)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 490)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 482)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 539)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 531)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 588)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 580)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 637)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 629)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 686)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 678)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 735)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && (1 <= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 727)] : 0.000000e+00f);
+          kernel_shared[(((int)threadIdx.x) * 4)] = kernel[(((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3))];
+          kernel_shared[((((int)threadIdx.x) * 4) + 1)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 9)];
+          kernel_shared[((((int)threadIdx.x) * 4) + 2)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 18)];
+          kernel_shared[((((int)threadIdx.x) * 4) + 3)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 27)];
+          if (((int)threadIdx.x) < 15) {
+            kernel_shared[((((int)threadIdx.x) * 4) + 196)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 4) & 15) * 9)) + (ry_outer_outer * 3))];
+            kernel_shared[((((int)threadIdx.x) * 4) + 197)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 5) & 15) * 9)) + (ry_outer_outer * 3))];
+            kernel_shared[((((int)threadIdx.x) * 4) + 198)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 6) & 15) * 9)) + (ry_outer_outer * 3))];
+            kernel_shared[((((int)threadIdx.x) * 4) + 199)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 7) & 15) * 9)) + (ry_outer_outer * 3))];
+          }
+          __syncthreads();
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[0]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[32]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[64]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[96]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[128]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[160]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[192]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[224]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[16]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[48]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[80]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[112]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[144]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[176]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[208]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[240]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[1]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[33]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[65]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[97]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[129]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[161]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[193]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[225]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[17]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[49]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[81]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[113]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[145]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[177]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[209]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[241]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[2]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[34]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[66]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[98]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[130]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[162]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[194]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[226]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[18]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[50]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[82]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[114]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[146]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[178]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[210]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[242]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[3]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[35]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[67]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[99]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[131]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[163]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[195]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[227]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[19]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[51]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[83]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[115]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[147]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[179]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[211]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[243]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[4]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[36]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[68]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[100]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[132]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[164]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[196]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[228]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[20]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[52]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[84]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[116]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[148]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[180]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[212]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[244]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[5]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[37]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[69]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[101]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[133]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[165]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[197]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[229]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[21]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[53]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[85]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[117]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[149]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[181]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[213]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[245]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[6]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[38]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[70]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[102]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[134]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[166]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[198]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[230]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[22]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[54]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[86]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[118]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[150]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[182]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[214]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[246]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[7]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[39]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[71]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[103]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[135]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[167]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[199]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[231]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[23]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[55]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[87]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[119]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[151]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[183]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[215]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[247]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[8]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[40]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[72]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[104]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[136]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[168]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[200]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[232]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[24]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[56]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[88]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[120]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[152]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[184]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[216]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[248]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[9]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[41]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[73]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[105]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[137]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[169]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[201]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[233]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[25]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[57]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[89]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[121]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[153]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[185]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[217]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[249]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[10]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[42]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[74]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[106]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[138]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[170]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[202]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[234]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[26]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[58]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[90]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[122]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[154]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[186]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[218]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[250]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[11]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[43]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[75]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[107]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[139]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[171]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[203]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[235]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[27]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[59]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[91]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[123]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[155]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[187]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[219]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[251]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[12]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[44]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[76]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[108]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[140]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[172]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[204]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[236]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[28]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[60]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[92]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[124]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[156]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[188]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[220]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[252]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[13]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[45]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[77]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[109]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[141]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[173]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[205]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[237]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[29]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[61]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[93]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[125]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[157]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[189]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[221]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[253]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[14]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[46]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[78]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[110]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[142]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[174]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[206]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[238]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[30]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[62]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[94]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[126]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[158]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[190]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[222]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[254]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[15]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[47]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[79]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[111]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[143]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[175]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[207]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[239]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[31]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[63]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[95]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[127]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[159]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[191]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[223]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[255]));
+          __syncthreads();
+          pad_temp_shared[((int)threadIdx.x)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) - 7)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 49)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 42)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 98)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 91)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 147)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 140)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 196)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 189)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 245)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 238)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 294)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 287)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 343)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 336)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 392)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 385)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 441)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 434)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 490)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 483)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 539)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 532)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 588)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 581)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 637)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 630)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 686)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 679)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 735)] = (((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 728)] : 0.000000e+00f);
+          kernel_shared[(((int)threadIdx.x) * 4)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 1)];
+          kernel_shared[((((int)threadIdx.x) * 4) + 1)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 10)];
+          kernel_shared[((((int)threadIdx.x) * 4) + 2)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 19)];
+          kernel_shared[((((int)threadIdx.x) * 4) + 3)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 28)];
+          if (((int)threadIdx.x) < 15) {
+            kernel_shared[((((int)threadIdx.x) * 4) + 196)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 4) & 15) * 9)) + (ry_outer_outer * 3)) + 1)];
+            kernel_shared[((((int)threadIdx.x) * 4) + 197)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 5) & 15) * 9)) + (ry_outer_outer * 3)) + 1)];
+            kernel_shared[((((int)threadIdx.x) * 4) + 198)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 6) & 15) * 9)) + (ry_outer_outer * 3)) + 1)];
+            kernel_shared[((((int)threadIdx.x) * 4) + 199)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 7) & 15) * 9)) + (ry_outer_outer * 3)) + 1)];
+          }
+          __syncthreads();
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[0]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[32]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[64]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[96]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[128]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[160]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[192]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[224]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[16]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[48]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[80]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[112]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[144]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[176]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[208]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[240]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[1]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[33]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[65]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[97]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[129]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[161]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[193]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[225]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[17]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[49]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[81]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[113]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[145]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[177]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[209]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[241]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[2]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[34]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[66]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[98]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[130]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[162]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[194]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[226]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[18]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[50]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[82]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[114]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[146]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[178]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[210]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[242]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[3]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[35]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[67]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[99]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[131]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[163]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[195]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[227]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[19]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[51]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[83]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[115]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[147]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[179]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[211]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[243]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[4]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[36]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[68]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[100]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[132]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[164]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[196]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[228]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[20]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[52]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[84]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[116]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[148]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[180]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[212]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[244]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[5]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[37]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[69]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[101]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[133]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[165]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[197]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[229]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[21]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[53]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[85]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[117]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[149]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[181]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[213]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[245]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[6]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[38]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[70]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[102]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[134]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[166]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[198]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[230]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[22]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[54]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[86]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[118]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[150]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[182]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[214]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[246]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[7]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[39]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[71]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[103]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[135]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[167]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[199]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[231]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[23]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[55]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[87]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[119]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[151]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[183]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[215]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[247]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[8]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[40]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[72]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[104]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[136]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[168]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[200]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[232]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[24]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[56]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[88]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[120]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[152]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[184]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[216]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[248]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[9]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[41]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[73]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[105]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[137]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[169]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[201]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[233]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[25]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[57]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[89]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[121]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[153]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[185]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[217]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[249]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[10]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[42]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[74]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[106]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[138]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[170]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[202]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[234]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[26]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[58]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[90]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[122]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[154]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[186]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[218]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[250]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[11]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[43]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[75]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[107]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[139]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[171]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[203]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[235]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[27]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[59]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[91]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[123]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[155]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[187]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[219]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[251]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[12]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[44]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[76]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[108]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[140]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[172]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[204]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[236]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[28]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[60]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[92]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[124]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[156]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[188]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[220]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[252]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[13]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[45]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[77]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[109]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[141]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[173]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[205]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[237]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[29]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[61]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[93]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[125]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[157]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[189]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[221]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[253]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[14]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[46]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[78]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[110]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[142]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[174]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[206]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[238]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[30]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[62]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[94]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[126]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[158]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[190]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[222]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[254]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[15]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[47]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[79]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[111]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[143]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[175]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[207]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[239]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[31]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[63]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[95]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[127]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[159]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[191]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[223]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[255]));
+          __syncthreads();
+          pad_temp_shared[((int)threadIdx.x)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) - 6)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 49)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 43)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 98)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 92)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 147)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 141)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 196)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 190)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 245)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 239)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 294)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 288)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 343)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 337)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 392)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 386)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 441)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 435)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 490)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 484)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 539)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 533)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 588)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 582)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 637)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 631)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 686)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 680)] : 0.000000e+00f);
+          pad_temp_shared[(((int)threadIdx.x) + 735)] = ((((1 <= ((((int)threadIdx.x) / 7) + ry_outer_outer)) && (((((int)threadIdx.x) / 7) + ry_outer_outer) < 8)) && ((((int)threadIdx.x) % 7) < 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 729)] : 0.000000e+00f);
+          kernel_shared[(((int)threadIdx.x) * 4)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 2)];
+          kernel_shared[((((int)threadIdx.x) * 4) + 1)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 11)];
+          kernel_shared[((((int)threadIdx.x) * 4) + 2)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 20)];
+          kernel_shared[((((int)threadIdx.x) * 4) + 3)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) & 3) * 36)) + (ry_outer_outer * 3)) + 29)];
+          if (((int)threadIdx.x) < 15) {
+            kernel_shared[((((int)threadIdx.x) * 4) + 196)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 4) & 15) * 9)) + (ry_outer_outer * 3)) + 2)];
+            kernel_shared[((((int)threadIdx.x) * 4) + 197)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 5) & 15) * 9)) + (ry_outer_outer * 3)) + 2)];
+            kernel_shared[((((int)threadIdx.x) * 4) + 198)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 6) & 15) * 9)) + (ry_outer_outer * 3)) + 2)];
+            kernel_shared[((((int)threadIdx.x) * 4) + 199)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) >> 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 7) & 15) * 9)) + (ry_outer_outer * 3)) + 2)];
+          }
+          __syncthreads();
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[0]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[32]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[64]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[96]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[128]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[160]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[192]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[224]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[16]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[48]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[80]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[112]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[144]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[176]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[208]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[240]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[1]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[33]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[65]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[97]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[129]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[161]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[193]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[225]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[17]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[49]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[81]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[113]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[145]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[177]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[209]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[241]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[2]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[34]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[66]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[98]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[130]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[162]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[194]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[226]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[18]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[50]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[82]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[114]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[146]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[178]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[210]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[242]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[3]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[35]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[67]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[99]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[131]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[163]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[195]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[227]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[19]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[51]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[83]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[115]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[147]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[179]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[211]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[243]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[4]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[36]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[68]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[100]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[132]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[164]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[196]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[228]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[20]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[52]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[84]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[116]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[148]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[180]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[212]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[244]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[5]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[37]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[69]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[101]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[133]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[165]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[197]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[229]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[21]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[53]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[85]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[117]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[149]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[181]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[213]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[245]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[6]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[38]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[70]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[102]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[134]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[166]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[198]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[230]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[22]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[54]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[86]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[118]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[150]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[182]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[214]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[246]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[7]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[39]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[71]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[103]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[135]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[167]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[199]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[231]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[23]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[55]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[87]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[119]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[151]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[183]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[215]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[247]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[8]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[40]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[72]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[104]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[136]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[168]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[200]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[232]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[24]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[56]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[88]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[120]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[152]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[184]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[216]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[248]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[9]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[41]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[73]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[105]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[137]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[169]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[201]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[233]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[25]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[57]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[89]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[121]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[153]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[185]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[217]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[249]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[10]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[42]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[74]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[106]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[138]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[170]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[202]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[234]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[26]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[58]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[90]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[122]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[154]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[186]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[218]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[250]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[11]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[43]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[75]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[107]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[139]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[171]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[203]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[235]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[27]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[59]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[91]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[123]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[155]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[187]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[219]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[251]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[12]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[44]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[76]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[108]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[140]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[172]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[204]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[236]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[28]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[60]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[92]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[124]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[156]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[188]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[220]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[252]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[13]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[45]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[77]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[109]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[141]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[173]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[205]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[237]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[29]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[61]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[93]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[125]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[157]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[189]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[221]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[253]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[14]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[46]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[78]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[110]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[142]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[174]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[206]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[238]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[30]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[62]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[94]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[126]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[158]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[190]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[222]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[254]));
+          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[15]));
+          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[47]));
+          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[79]));
+          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[111]));
+          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[143]));
+          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[175]));
+          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[207]));
+          conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[239]));
+          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[31]));
+          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[63]));
+          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[95]));
+          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[127]));
+          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[159]));
+          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[191]));
+          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[223]));
+          conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[255]));
         }
-        __syncthreads();
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[0]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[9]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[36]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[45]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[72]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[81]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[108]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[117]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[1]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[10]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[37]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[46]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[73]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[82]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[109]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[118]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[2]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[11]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[38]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[47]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[74]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[83]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[110]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[119]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[3]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[12]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[39]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[48]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[75]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[84]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[111]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[120]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[4]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[13]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[40]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[49]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[76]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[85]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[112]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[121]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[5]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[14]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[41]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[50]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[77]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[86]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[113]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[122]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[6]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[15]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[42]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[51]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[78]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[87]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[114]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[123]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[7]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[16]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[43]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[52]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[79]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[88]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[115]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[124]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[8]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[17]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[44]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[53]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[80]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[89]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[116]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[125]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[18]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[27]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[54]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[63]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[90]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[99]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[126]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[135]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[19]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[28]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[55]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[64]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[91]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[100]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[127]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[136]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[20]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[29]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[56]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[65]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[92]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[101]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[128]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[137]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[21]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[30]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[57]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[66]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[93]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[102]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[129]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[138]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[22]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[31]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[58]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[67]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[94]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[103]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[130]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[139]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[23]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[32]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[59]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[68]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[95]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[104]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[131]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[140]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[24]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[33]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[60]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[69]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[96]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[105]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[132]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[141]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[25]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[34]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[61]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[70]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[97]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[106]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[133]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[142]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[26]));
-        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[35]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[62]));
-        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[71]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[98]));
-        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[107]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[134]));
-        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[143]));
       }
-      for (int i1_inner = 0; i1_inner < 4; ++i1_inner) {
-        compute[(((((int)blockIdx.x) * 196) + (i1_inner * 49)) + ((int)threadIdx.x))] = max((conv2d_nchw[i1_inner] + bias[((((int)blockIdx.x) * 4) + i1_inner)]), 0.000000e+00f);
+      for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
+        compute[(((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x))] = max((conv2d_nchw[i1_inner] + bias[((((int)blockIdx.x) * 16) + i1_inner)]), 0.000000e+00f);
+        compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 98)] = max((conv2d_nchw[(i1_inner + 2)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 2)]), 0.000000e+00f);
+        compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 196)] = max((conv2d_nchw[(i1_inner + 4)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 4)]), 0.000000e+00f);
+        compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 294)] = max((conv2d_nchw[(i1_inner + 6)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 6)]), 0.000000e+00f);
+        compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 392)] = max((conv2d_nchw[(i1_inner + 8)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 8)]), 0.000000e+00f);
+        compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 490)] = max((conv2d_nchw[(i1_inner + 10)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 10)]), 0.000000e+00f);
+        compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 588)] = max((conv2d_nchw[(i1_inner + 12)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 12)]), 0.000000e+00f);
+        compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 686)] = max((conv2d_nchw[(i1_inner + 14)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 14)]), 0.000000e+00f);
       }
     }
 
@@ -825,7 +2299,7 @@ In the example below we resume the status and do more 5 trials.
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/xgboost/training.py:17: UserWarning: Old style callback is deprecated.  See: https://xgboost.readthedocs.io/en/latest/python/callbacks.html
       warnings.warn(f'Old style callback is deprecated.  See: {link}', UserWarning)
     Get devices for measurement successfully!
-    .T
+
 
 
 
@@ -833,7 +2307,7 @@ 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  52.239 seconds)
+   **Total running time of the script:** ( 5 minutes  43.754 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 175fd78ad6..86cdd9f6e4 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
@@ -643,7 +643,7 @@ so we can read the log file and load the best schedules.
     Evaluate inference time cost...
     Execution time summary:
      mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)  
-       8.2691       8.2682       8.2786       8.2605       0.0074   
+       8.2044       8.2046       8.2096       8.1990       0.0043   
                
 
 
@@ -671,7 +671,7 @@ Other Tips
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  4.373 seconds)
+   **Total running time of the script:** ( 1 minutes  4.752 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 ca33a85f14..631d6a07db 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
@@ -662,7 +662,7 @@ 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)  
-      765.1622     765.7585     766.9518     762.7763      1.7560   
+      763.1550     763.9276     764.1947     761.3427      1.2861   
                
 
 
@@ -690,7 +690,7 @@ Other Tips
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  34.879 seconds)
+   **Total running time of the script:** ( 1 minutes  34.813 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 55f0ff3cad..eae3ac0f0b 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
@@ -386,30 +386,76 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
                  placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [128, 512], []),
                  compute: Buffer(compute_2: Pointer(float32), float32, [128, 512], [])}
       buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute} {
-      for (i0.outer.i1.outer.fused: int32, 0, 128) "parallel" {
-        allocate(compute_3: Pointer(global float32), float32, [512]), storage_scope = global {
-          for (i.outer.inner: int32, 0, 8) {
-            for (i.inner.init: int32, 0, 4) {
-              for (j.init: int32, 0, 16) {
-                compute_4: Buffer(compute_3, float32, [512], [])[(((i.outer.inner*64) + (i.inner.init*16)) + j.init)] = 0f32
+      for (i0.outer.i1.outer.fused: int32, 0, 256) "parallel" {
+        allocate(compute_3: Pointer(global float32), float32, [256]), storage_scope = global {
+          for (nb_j.inner: int32, 0, 2) {
+            for (i.inner.init: int32, 0, 8) {
+              let cse_var_1: int32 = ((i.inner.init*32) + (nb_j.inner*16))
+               {
+                compute_4: Buffer(compute_3, float32, [256], [])[cse_var_1] = 0f32
+                compute_4[(cse_var_1 + 1)] = 0f32
+                compute_4[(cse_var_1 + 2)] = 0f32
+                compute_4[(cse_var_1 + 3)] = 0f32
+                compute_4[(cse_var_1 + 4)] = 0f32
+                compute_4[(cse_var_1 + 5)] = 0f32
+                compute_4[(cse_var_1 + 6)] = 0f32
+                compute_4[(cse_var_1 + 7)] = 0f32
+                compute_4[(cse_var_1 + 8)] = 0f32
+                compute_4[(cse_var_1 + 9)] = 0f32
+                compute_4[(cse_var_1 + 10)] = 0f32
+                compute_4[(cse_var_1 + 11)] = 0f32
+                compute_4[(cse_var_1 + 12)] = 0f32
+                compute_4[(cse_var_1 + 13)] = 0f32
+                compute_4[(cse_var_1 + 14)] = 0f32
+                compute_4[(cse_var_1 + 15)] = 0f32
               }
             }
-            for (elem_idx: int32, 0, let cse_var_1: int32 = floormod(i0.outer.i1.outer.fused, 32) in (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(cse_var_1 + 1)] - placeholder_15[cse_var_1])) {
-              for (i.inner: int32, 0, 4) {
-                for (j: int32, 0, 16) {
-                  let cse_var_2: int32 = floormod(i0.outer.i1.outer.fused, 32)
-                  if @tir.likely((elem_idx < (placeholder_15[(cse_var_2 + 1)] - placeholder_15[cse_var_2])), dtype=bool) {
-                    let cse_var_3: int32 = (((i.outer.inner*64) + (i.inner*16)) + j)
-                    compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[(((placeholder_15[cse_var_2]*16) + (elem_idx*16)) + j)]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[((((floordiv(i0.outer.i1.outer.fused, 32)*8192) + (i.outer.inner*1024)) + (i.inner*256)) + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[cse_var_2] + elem_idx)])], 0f32)))
-                  }
+            for (elem_idx: int32, 0, let cse_var_2: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner) in (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(cse_var_2 + 1)] - placeholder_15[cse_var_2])) {
+              for (i.inner: int32, 0, 8) {
+                let cse_var_21: int32 = (elem_idx*16)
+                let cse_var_20: int32 = ((i.inner*32) + (nb_j.inner*16))
+                let cse_var_19: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner)
+                let cse_var_18: int32 = ((floordiv(i0.outer.i1.outer.fused, 16)*2048) + (i.inner*256))
+                let cse_var_17: int32 = (cse_var_20 + 9)
+                let cse_var_16: int32 = (cse_var_20 + 8)
+                let cse_var_15: int32 = (cse_var_20 + 7)
+                let cse_var_14: int32 = (cse_var_20 + 6)
+                let cse_var_13: int32 = (cse_var_20 + 5)
+                let cse_var_12: int32 = (cse_var_20 + 4)
+                let cse_var_11: int32 = (cse_var_20 + 3)
+                let cse_var_10: int32 = (cse_var_20 + 2)
+                let cse_var_9: int32 = (cse_var_20 + 15)
+                let cse_var_8: int32 = (cse_var_20 + 14)
+                let cse_var_7: int32 = (cse_var_20 + 13)
+                let cse_var_6: int32 = (cse_var_20 + 12)
+                let cse_var_5: int32 = (cse_var_20 + 11)
+                let cse_var_4: int32 = (cse_var_20 + 10)
+                let cse_var_3: int32 = (cse_var_20 + 1)
+                 {
+                  compute_4[cse_var_20] = (compute_4[cse_var_20] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[((placeholder_15[cse_var_19]*16) + cse_var_21)]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[(cse_var_18 + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 1)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_10] = (compute_4[cse_var_10] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 2)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_11] = (compute_4[cse_var_11] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 3)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_12] = (compute_4[cse_var_12] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 4)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_13] = (compute_4[cse_var_13] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 5)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_14] = (compute_4[cse_var_14] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 6)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_15] = (compute_4[cse_var_15] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 7)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_16] = (compute_4[cse_var_16] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 8)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_17] = (compute_4[cse_var_17] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 9)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_4] = (compute_4[cse_var_4] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 10)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_5] = (compute_4[cse_var_5] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 11)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_6] = (compute_4[cse_var_6] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 12)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_7] = (compute_4[cse_var_7] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 13)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_8] = (compute_4[cse_var_8] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 14)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+                  compute_4[cse_var_9] = (compute_4[cse_var_9] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 15)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
                 }
               }
             }
           }
-          for (i0.inner: int32, 0, 32) {
-            for (i1.inner: int32, 0, 16) {
-              let cse_var_4: int32 = ((((floordiv(i0.outer.i1.outer.fused, 32)*16384) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 32)*16)) + i1.inner)
-              compute_5: Buffer(compute_2, float32, [65536], [])[cse_var_4] = max((compute_4[((i0.inner*16) + i1.inner)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[cse_var_4]), 0f32)
+          for (i0.inner: int32, 0, 8) {
+            for (i1.inner: int32, 0, 32) {
+              let cse_var_22: int32 = ((((floordiv(i0.outer.i1.outer.fused, 16)*4096) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 16)*32)) + i1.inner)
+              compute_5: Buffer(compute_2, float32, [65536], [])[cse_var_22] = max((compute_4[((i0.inner*32) + i1.inner)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[cse_var_22]), 0f32)
             }
           }
         }
@@ -466,7 +512,7 @@ We build the binary and check its correctness and performance.
 
  .. code-block:: none
 
-    Execution time of this operator: 1.477 ms
+    Execution time of this operator: 1.906 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 57c9bc35a2..25f8446b3e 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
@@ -5,12 +5,12 @@
 
 Computation times
 =================
-**00:31.477** total execution time for **how_to_tune_with_autotvm** files:
+**00:35.755** 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``)           | 00:31.441 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``)           | 00:35.720 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``)               | 00:00.021 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``)               | 00:00.019 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_cuda.py` (``tune_relay_cuda.py``)             | 00:00.005 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
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 f4ce89995b..948d1f6800 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
@@ -265,7 +265,8 @@ for this template
     waiting for device...
     device available
     Get devices for measurement successfully!
-    No: 1   GFLOPS: 0.00/0.00       result: Traceback (most recent call last):
+    No: 1   GFLOPS: 2.32/2.32       result: MeasureResult(costs=(0.099728897,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.013670921325684, timestamp=1668620461.9634821) [('tile_f', [-1, 8, 8, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,1781024
+    No: 2   GFLOPS: 0.00/2.32       result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -387,8 +388,9 @@ 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 871, 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, 32, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3647031
-    No: 2   GFLOPS: 0.00/0.00       result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 16, 1, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10440020
+    No: 3   GFLOPS: 1.98/2.32       result: MeasureResult(costs=(0.11707182299999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.8137314319610596, timestamp=1668620464.7798214)        [('tile_f', [-1, 128, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2561687
+    No: 4   GFLOPS: 0.00/2.32       result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -510,9 +512,8 @@ 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 871, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 8, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5803149
-    No: 3   GFLOPS: 228.85/228.85   result: MeasureResult(costs=(0.0010115684646464648,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.749254941940308, timestamp=1668618422.8077989)       [('tile_f', [-1, 4, 2, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7984692
-    No: 4   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 64, 1, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 256, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3355501
+    No: 5   GFLOPS: 0.00/2.32       result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -634,8 +635,8 @@ 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 871, 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, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,350595
-    No: 5   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 2, 32]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3036631
+    No: 6   GFLOPS: 0.00/2.32       result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -757,9 +758,9 @@ 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 871, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 16, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9650018
-    No: 6   GFLOPS: 126.37/228.85   result: MeasureResult(costs=(0.001831866727272727,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.34965181350708, timestamp=1668618425.2790372) [('tile_f', [-1, 1, 4, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2177455
-    No: 7   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 8, 8, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 512, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7778477
+    No: 7   GFLOPS: 10.83/10.83     result: MeasureResult(costs=(0.021368804833333335,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6941542625427246, timestamp=1668620468.6158936)       [('tile_f', [-1, 32, 8, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3326432
+    No: 8   GFLOPS: 0.00/10.83      result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -881,8 +882,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 871, 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, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7101279
-    No: 8   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 256, 2, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 16, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6304998
+    No: 9   GFLOPS: 0.00/10.83      result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1004,8 +1005,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 871, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 128, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5089138
-    No: 9   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 64, 4, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 256, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,609645
+    No: 10  GFLOPS: 0.00/10.83      result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1127,8 +1128,10 @@ 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 871, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 32, 4, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,943480
-    No: 10  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 4, 4, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9569677
+    No: 11  GFLOPS: 160.31/160.31   result: MeasureResult(costs=(0.0014441192545454544,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.183053970336914, timestamp=1668620471.017138)        [('tile_f', [-1, 2, 4, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9412773
+    No: 12  GFLOPS: 15.38/160.31    result: MeasureResult(costs=(0.015050043285714285,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.1918892860412598, timestamp=1668620471.784751)        [('tile_f', [-1, 8, 16, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2520449
+    No: 13  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1250,8 +1253,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 871, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 4, 64, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7111376
-    No: 11  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 8, 2, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4276066
+    No: 14  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1373,8 +1376,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 871, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 128, 2, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8769711
-    No: 12  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 8, 4, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3255932
+    No: 15  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1496,8 +1499,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 871, 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, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 32, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8437575
-    No: 13  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 8, 32, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3803233
+    No: 16  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1619,8 +1622,9 @@ 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 871, 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, 2, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6925666
-    No: 14  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 8, 1, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5578288
+    No: 17  GFLOPS: 36.23/160.31    result: MeasureResult(costs=(0.006389735090909091,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.769231081008911, timestamp=1668620476.9480226)        [('tile_f', [-1, 1, 4, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4576115
+    No: 18  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1742,8 +1746,8 @@ 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 871, 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, 1, 512]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2489299
-    No: 15  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 4, 1, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7713998
+    No: 19  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1865,8 +1869,8 @@ 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 871, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 256, 2, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2867938
-    No: 16  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 2, 32]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8664231
+    No: 20  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
@@ -1988,133 +1992,7 @@ 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 871, 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, 256, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9065199
-    No: 17  GFLOPS: 53.80/228.85    result: MeasureResult(costs=(0.00430309384,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4398438930511475, timestamp=1668618429.2793698)      [('tile_f', [-1, 1, 64, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6780343
-    No: 18  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
-      File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 588, in __call__
-        func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
-      File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 540, in _build_func_common
-        func = build(s, args, target_host=task.target_host, runtime=runtime)
-      File "/workspace/python/tvm/driver/build_module.py", line 227, in build
-        input_mod = lower(inputs, args, name=name, binds=binds)
-      File "/workspace/python/tvm/driver/build_module.py", line 134, in lower
-        return ffi.lower_schedule(inp, args, name, binds, simple_mode)
-      File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
-      File "tvm/_ffi/_cython/./packed_func.pxi", line 276, in tvm._ffi._cy3.core.FuncCall
-      File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
-    tvm._ffi.base.TVMError: Traceback (most recent call last):
-      24: TVMFuncCall
-            at ../src/runtime/c_runtime_api.cc:477
-      23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
-            at ../include/tvm/runtime/packed_func.h:1217
-      22: Call
-            at ../include/tvm/runtime/packed_func.h:1213
-      21: operator()
-            at ../include/tvm/runtime/packed_func.h:1731
-      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:1671
-      19: run<>
-            at ../include/tvm/runtime/packed_func.h:1631
-      18: run<tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1631
-      17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1631
-      16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1631
-      15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1631
-      14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1646
-      13: operator()
-            at ../src/driver/driver_api.cc:388
-      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:374
-      11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
-            at ../src/driver/driver_api.cc:269
-      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:453
-      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:1750
-      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:1694
-      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:1618
-      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 871, 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:1731
-      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:1671
-      19: run<>
-            at ../include/tvm/runtime/packed_func.h:1631
-      18: run<tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1631
-      17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1631
-      16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1631
-      15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1631
-      14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
-            at ../include/tvm/runtime/packed_func.h:1646
-      13: operator()
-            at ../src/driver/driver_api.cc:388
-      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:374
-      11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
-            at ../src/driver/driver_api.cc:269
-      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:453
-      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:1750
-      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:1694
-      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:1618
-      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 871, in verify_pass
-        raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 4, 4, 16]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,639717
-    No: 19  GFLOPS: 37.57/228.85    result: MeasureResult(costs=(0.006161984117647059,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.3533060550689697, timestamp=1668618430.00153) [('tile_f', [-1, 32, 2, 1]), ('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', 1)],None,5658415
-    No: 20  GFLOPS: 2.70/228.85     result: MeasureResult(costs=(0.085883934,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.988248109817505, timestamp=1668618431.4606593) [('tile_f', [-1, 32, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2554875
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 256, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,943632
 
 
 
@@ -2169,9 +2047,9 @@ and measure running time.
     Finish loading 20 records
 
     Best config:
-    [('tile_f', [-1, 4, 2, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7984692
+    [('tile_f', [-1, 2, 4, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9412773
     Finish loading 20 records
-    Time cost of this operator: 0.000809
+    Time cost of this operator: 0.001880
 
 
 
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 924b1c7bc1..9a0fc4929c 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
@@ -327,10 +327,10 @@ 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  313.9     98.712   (1, 2, 10, 10, 3)  2       1        [313.9]           
-    tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       3.131     0.985    (1, 6, 10, 10)     1       1        [3.131]           
-    tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.964     0.303    (1, 1, 10, 10, 3)  1       1        [0.964]           
-    Total_time                                    -                                             317.996   -        -                  -       -        -                 
+    tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  309.4     98.728   (1, 2, 10, 10, 3)  2       1        [309.4]           
+    tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       3.023     0.965    (1, 6, 10, 10)     1       1        [3.023]           
+    tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.962     0.307    (1, 1, 10, 10, 3)  1       1        [0.962]           
+    Total_time                                    -                                             313.386   -        -                  -       -        -                 
 
 
 
@@ -394,10 +394,10 @@ Timing the tuned program
     ########## Build with Autotuning ##########
     Node Name                                     Ops                                           Time(us)  Time(%)  Shape              Inputs  Outputs  Measurements(us)  
     ---------                                     ---                                           --------  -------  -----              ------  -------  ----------------  
-    tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  102.6     97.476   (1, 6, 10, 10, 1)  2       1        [102.6]           
-    tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       1.799     1.709    (1, 6, 10, 10)     1       1        [1.799]           
-    tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.858     0.815    (1, 3, 10, 10, 1)  1       1        [0.858]           
-    Total_time                                    -                                             105.257   -        -                  -       -        -                 
+    tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  101.8     97.407   (1, 6, 10, 10, 1)  2       1        [101.8]           
+    tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       1.756     1.68     (1, 6, 10, 10)     1       1        [1.756]           
+    tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.955     0.913    (1, 1, 10, 10, 3)  1       1        [0.955]           
+    Total_time                                    -                                             104.51    -        -                  -       -        -                 
 
 
 
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 814f6e0438..9da436b52b 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
@@ -109,7 +109,7 @@ 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, 75.6MB/s]
+
      0%|          | 0.00/3.42M [00:00<?, ?B/s]
     76%|#######6  | 2.61M/3.42M [00:00<00:00, 27.3MB/s]
    100%|##########| 3.42M/3.42M [00:00<00:00, 27.9MB/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.
@@ -314,7 +314,7 @@ Look up prediction top 1 index in 1000 class synset.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  4.241 seconds)
+   **Total running time of the script:** ( 1 minutes  5.341 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 d4ec269a98..594787074d 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
@@ -225,7 +225,7 @@ take about **2 minutes** to download the Stanford Cars, while COCO 2017 validati
  .. code-block:: none
 
 
-    '/tmp/tmpbz6hkicy/images/random'
+    '/tmp/tmpe0nb0235/images/random'
 
 
 
@@ -325,8 +325,8 @@ objects to other stuff? We can display some examples from our datasets using ``m
 
  .. code-block:: none
 
-    /tmp/tmpbz6hkicy/images/target contains 8144 images
-    /tmp/tmpbz6hkicy/images/random contains 5000 images
+    /tmp/tmpe0nb0235/images/target contains 8144 images
+    /tmp/tmpe0nb0235/images/random contains 5000 images
 
 
 
@@ -501,13 +501,13 @@ the time on our validation set).
  .. code-block:: none
 
     Epoch 1/3
-    328/328 - 47s - loss: 0.2374 - accuracy: 0.9209 - val_loss: 0.1806 - val_accuracy: 0.9388 - 47s/epoch - 143ms/step
+    328/328 - 47s - loss: 0.2109 - accuracy: 0.9260 - val_loss: 0.1198 - val_accuracy: 0.9569 - 47s/epoch - 144ms/step
     Epoch 2/3
-    328/328 - 44s - loss: 0.1106 - accuracy: 0.9570 - val_loss: 0.0969 - val_accuracy: 0.9641 - 44s/epoch - 133ms/step
+    328/328 - 44s - loss: 0.1042 - accuracy: 0.9624 - val_loss: 0.1374 - val_accuracy: 0.9498 - 44s/epoch - 133ms/step
     Epoch 3/3
-    328/328 - 43s - loss: 0.0756 - accuracy: 0.9724 - val_loss: 0.0981 - val_accuracy: 0.9690 - 43s/epoch - 132ms/step
+    328/328 - 44s - loss: 0.0650 - accuracy: 0.9738 - val_loss: 0.1059 - val_accuracy: 0.9660 - 44s/epoch - 133ms/step
 
-    <keras.callbacks.History object at 0x7f5cc00bee90>
+    <keras.callbacks.History object at 0x7f34dd15a3d0>
 
 
 
@@ -864,7 +864,7 @@ Arduino tutorial for how to do that `on GitHub <https://github.com/guberti/tvm-a
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 4 minutes  47.987 seconds)
+   **Total running time of the script:** ( 4 minutes  42.129 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 80c36ae27e..e1265c5993 100644
--- a/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
@@ -5,18 +5,18 @@
 
 Computation times
 =================
-**06:54.261** total execution time for **how_to_work_with_microtvm** files:
+**06:51.708** total execution time for **how_to_work_with_microtvm** files:
 
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``)               | 04:47.987 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``)               | 04:42.129 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``)           | 01:04.241 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``)           | 01:05.341 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``)         | 00:49.982 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``)         | 00:51.638 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``)                   | 00:08.264 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``)                   | 00:08.628 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``)             | 00:03.784 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``)             | 00:03.969 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_work_with_microtvm_micro_reference_vm.py` (``micro_reference_vm.py``) | 00:00.001 | 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 f7a06f5e88..2aca469e19 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
 =================
-**00:43.314** total execution time for **how_to_work_with_relay** files:
+**00:44.784** 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``) | 00:31.790 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:32.784 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``)           | 00:10.146 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``)           | 00:10.272 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``)                             | 00:01.372 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``)                             | 00:01.721 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``)                 | 00:00.007 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt b/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
index 4a1db0750a..83ed44ba4f 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
@@ -261,7 +261,7 @@ The following example customizes CUDA lowering rule for :code:`exp`.
  .. code-block:: none
 
 
-    <function my_cuda_math_rule at 0x7f5cc13808c0>
+    <function my_cuda_math_rule at 0x7f330c705e60>
 
 
 
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 9f6e90c7ad..db4b13a3c6 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
 =================
-**00:07.373** total execution time for **how_to_work_with_schedules** files:
+**00:06.448** total execution time for **how_to_work_with_schedules** files:
 
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``)                 | 00:05.026 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``)                 | 00:03.974 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``)                     | 00:01.043 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``)                     | 00:01.086 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``)                     | 00:00.555 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``)                     | 00:00.594 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``)                               | 00:00.535 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``)                               | 00:00.577 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``)                     | 00:00.116 | 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_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.049 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.050 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``)                               | 00:00.029 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``)                               | 00:00.030 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``)               | 00:00.019 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``)               | 00:00.020 | 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 561b2f532f..3d98da4987 100644
--- a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
@@ -343,7 +343,7 @@ The importing needs to happen before the tensorized GEMV being executed.
                  B: Buffer(B_2: Pointer(float32), float32, [512, 64], []),
                  C: Buffer(C_2: Pointer(float32), float32, [1024, 512], [])}
       buffer_map = {A_1: A, B_1: B, C_1: C} {
-      attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmp0b8qkn1u/input0.cc'\nsource_filename = \"/tmp/tmp0b8qkn1u/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n  %7 = alloca float*, align 8\n  %8 = alloca float*, align 8\n  %9 = alloca floa [...]
+      attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmp8j_p51zl/input0.cc'\nsource_filename = \"/tmp/tmp8j_p51zl/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n  %7 = alloca float*, align 8\n  %8 = alloca float*, align 8\n  %9 = alloca floa [...]
       for (i, 0, 1024) {
         for (j.outer: int32, 0, 32) {
           @tir.call_extern("gemv_update", @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), C_2, ((i*512) + (j.outer*16)), 16, 2, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), A_2, (i*64), 64, 1, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), B_2, (j.outer*1024), 1024, 1, dtype=handle), 16, 64, 64, dtype=int32)
diff --git a/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt
index 671b4b647c..a0018ca40d 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
 =================
-**00:26.589** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:27.412** total execution time for **topic_vta_tutorials_autotvm** files:
 
 +---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:26.582 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:27.406 | 0.0 MB |
 +---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``)     | 00:00.006 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``)     | 00:00.007 | 0.0 MB |
 +---------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
index 859c2ce2c9..2731db312e 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
@@ -289,7 +289,7 @@ 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 29.82s!
+    resnet18_v1 inference graph built in 30.35s!
 
 
 
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 ec6e486667..582dbfbde3 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
@@ -333,7 +333,7 @@ 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 20.23s!
+    yolov3-tiny inference graph built in 20.43s!
 
 
 
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 aa1a74a8fe..6f102ab2a3 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
 =================
-**01:42.683** total execution time for **topic_vta_tutorials_frontend** files:
+**01:43.048** total execution time for **topic_vta_tutorials_frontend** files:
 
 +------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``)           | 00:52.995 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``)           | 00:52.739 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:49.688 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:50.309 | 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 c85d4edcce..1db2181846 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
 =================
-**00:03.153** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.235** total execution time for **topic_vta_tutorials_optimize** files:
 
 +--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``)         | 00:02.714 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``)         | 00:02.763 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.440 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.472 | 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 917c9643bb..bf6716f4f4 100644
--- a/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
 
 Computation times
 =================
-**00:00.764** total execution time for **topic_vta_tutorials** files:
+**00:00.837** total execution time for **topic_vta_tutorials** files:
 
 +---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.408 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.445 | 0.0 MB |
 +---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.357 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.392 | 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 a24990e7c8..1e6e2c2f4a 100644
--- a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
@@ -325,7 +325,7 @@ We build the binary and check its correctness and performance.
 
  .. code-block:: none
 
-    Execution time of this operator: 93.327 ms
+    Execution time of this operator: 100.349 ms
 
 
 
@@ -443,7 +443,7 @@ operations.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  16.007 seconds)
+   **Total running time of the script:** ( 1 minutes  21.583 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 b48792c4c7..b7f1e492b6 100644
--- a/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
@@ -450,16 +450,16 @@ reduce variance, we take 5 measurements and average them.
     waiting for device...
     device available
     Get devices for measurement successfully!
-    No: 1   GFLOPS: 11.03/11.03     result: MeasureResult(costs=(0.024329743600000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5313427448272705, timestamp=1668617045.4020307)       [('tile_y', [-1, 1]), ('tile_x', [-1, 512])],None,90
-    No: 2   GFLOPS: 12.24/12.24     result: MeasureResult(costs=(0.0219322012,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5325038433074951, timestamp=1668617046.6605668)       [('tile_y', [-1, 8]), ('tile_x', [-1, 256])],None,83
-    No: 3   GFLOPS: 14.66/14.66     result: MeasureResult(costs=(0.0183090816,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.7388677597045898, timestamp=1668617047.8500664)       [('tile_y', [-1, 64]), ('tile_x', [-1, 64])],None,66
-    No: 4   GFLOPS: 12.39/14.66     result: MeasureResult(costs=(0.021670452599999997,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.4895329475402832, timestamp=1668617048.365852)        [('tile_y', [-1, 256]), ('tile_x', [-1, 512])],None,98
-    No: 5   GFLOPS: 8.70/14.66      result: MeasureResult(costs=(0.030850067400000004,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.8019249439239502, timestamp=1668617049.2892187)       [('tile_y', [-1, 16]), ('tile_x', [-1, 64])],None,64
-    No: 6   GFLOPS: 12.55/14.66     result: MeasureResult(costs=(0.021383448000000003,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5027682781219482, timestamp=1668617049.800311)        [('tile_y', [-1, 256]), ('tile_x', [-1, 256])],None,88
-    No: 7   GFLOPS: 12.11/14.66     result: MeasureResult(costs=(0.0221694228,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.593656063079834, timestamp=1668617051.070341) [('tile_y', [-1, 1]), ('tile_x', [-1, 64])],None,60
-    No: 8   GFLOPS: 4.03/14.66      result: MeasureResult(costs=(0.0666180478,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2662608623504639, timestamp=1668617052.3394432)       [('tile_y', [-1, 4]), ('tile_x', [-1, 16])],None,42
-    No: 9   GFLOPS: 13.09/14.66     result: MeasureResult(costs=(0.0205115432,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.4462544918060303, timestamp=1668617052.900465)        [('tile_y', [-1, 4]), ('tile_x', [-1, 512])],None,92
-    No: 10  GFLOPS: 1.71/14.66      result: MeasureResult(costs=(0.15705507219999998,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.621229410171509, timestamp=1668617055.5771248) [('tile_y', [-1, 512]), ('tile_x', [-1, 4])],None,29
+    No: 1   GFLOPS: 2.59/2.59       result: MeasureResult(costs=(0.10359291479999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8303396701812744, timestamp=1668619046.14139)  [('tile_y', [-1, 8]), ('tile_x', [-1, 4])],None,23
+    No: 2   GFLOPS: 0.86/2.59       result: MeasureResult(costs=(0.31212432100000004,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.175588369369507, timestamp=1668619051.3432844) [('tile_y', [-1, 512]), ('tile_x', [-1, 2])],None,19
+    No: 3   GFLOPS: 12.48/12.48     result: MeasureResult(costs=(0.021505876200000003,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.48230600357055664, timestamp=1668619052.6419113)      [('tile_y', [-1, 128]), ('tile_x', [-1, 512])],None,97
+    No: 4   GFLOPS: 11.90/12.48     result: MeasureResult(costs=(0.0225636806,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5249507427215576, timestamp=1668619053.9455721)       [('tile_y', [-1, 256]), ('tile_x', [-1, 256])],None,88
+    No: 5   GFLOPS: 1.08/12.48      result: MeasureResult(costs=(0.24850028739999996,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.030014753341675, timestamp=1668619058.1830983) [('tile_y', [-1, 16]), ('tile_x', [-1, 1])],None,4
+    No: 6   GFLOPS: 1.63/12.48      result: MeasureResult(costs=(0.16475953640000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.791689157485962, timestamp=1668619061.7600245) [('tile_y', [-1, 64]), ('tile_x', [-1, 4])],None,26
+    No: 7   GFLOPS: 10.96/12.48     result: MeasureResult(costs=(0.0245022136,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6264443397521973, timestamp=1668619062.3191602)       [('tile_y', [-1, 1]), ('tile_x', [-1, 64])],None,60
+    No: 8   GFLOPS: 10.91/12.48     result: MeasureResult(costs=(0.0245933216,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5864930152893066, timestamp=1668619062.924451)        [('tile_y', [-1, 2]), ('tile_x', [-1, 256])],None,81
+    No: 9   GFLOPS: 4.13/12.48      result: MeasureResult(costs=(0.065029372,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.1726446151733398, timestamp=1668619064.210647) [('tile_y', [-1, 16]), ('tile_x', [-1, 16])],None,44
+    No: 10  GFLOPS: 2.12/12.48      result: MeasureResult(costs=(0.1264244674,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.1329433917999268, timestamp=1668619066.3955705)       [('tile_y', [-1, 512]), ('tile_x', [-1, 8])],None,39
 
 
 
diff --git a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
index f161addcad..c8075c8422 100644
--- a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
@@ -320,7 +320,7 @@ standard deviation.
 
  .. code-block:: none
 
-    {'mean': 516.8028887700029, 'median': 516.7933151499938, 'std': 1.8442027952596247}
+    {'mean': 521.135307909999, 'median': 521.1086226000134, 'std': 2.045413370611934}
 
 
 
@@ -554,30 +554,32 @@ the tuning data to.
 
  .. code-block:: none
 
-
    [Task  1/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  1/25]  Current/Best:   14.83/  17.52 GFLOPS | Progress: (4/20) | 6.94 s
    [Task  1/25]  Current/Best:   22.90/  22.90 GFLOPS | Progress: (8/20) | 13.12 s
    [Task  1/25]  Current/Best:    5.64/  22.90 GFLOPS | Progress: (12/20) | 16.27 s
    [Task  1/25]  Current/Best:    5.76/  22.90 GFLOPS | Progress: (16/20) | 19.33 s
    [Task  1/25]  Current/Best:    5.27/  22.90 GFLOPS | Progress: (20/20) | 23.12 s Done.
-
    [Task  2/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  2/25]  Current/Best:   21.38/  21.38 GFLOPS | Progress: (4/20) | 2.70 s
    [Task  2/25]  Current/Best:   13.71/  21.38 GFLOPS | Progress: (8/20) | 4.44 s
    [Task  2/25]  Current/Best:    7.99/  21.38 GFLOPS | Progress: (12/20) | 5.51 s
    [Task  2/25]  Current/Best:   13.04/  21.38 GFLOPS | Progress: (16/20) | 7.34 s
    [Task  2/25]  Current/Best:   12.92/  21.38 GFLOPS | Progress: (20/20) | 9.32 s Done.
-
    [Task  3/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  3/25]  Current/Best:   10.70/  21.14 GFLOPS | Progress: (4/20) | 3.45 s
    [Task  3/25]  Current/Best:    6.22/  21.14 GFLOPS | Progress: (8/20) | 6.03 s
    [Task  3/25]  Current/Best:   15.77/  24.11 GFLOPS | Progress: (12/20) | 7.70 s
    [Task  3/25]  Current/Best:    7.64/  24.11 GFLOPS | Progress: (16/20) | 10.06 s
    [Task  3/25]  Current/Best:    5.62/  24.11 GFLOPS | Progress: (20/20) | 12.75 s Done.
-
    [Task  4/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  4/25]  Current/Best:   16.29/  16.29 GFLOPS | Progress: (4/20) | 6.83 s
    [Task  4/25]  Current/Best:    7.81/  16.82 GFLOPS | Progress: (8/20) | 9.05 s
    [Task  4/25]  Current/Best:   11.80/  16.82 GFLOPS | Progress: (12/20) | 10.93 s
    [Task  4/25]  Current/Best:    8.10/  20.87 GFLOPS | Progress: (16/20) | 12.89 s
    [Task  4/25]  Current/Best:   10.30/  20.87 GFLOPS | Progress: (20/20) | 14.43 s Done.
-
    [Task  5/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  5/25]  Current/Best:   16.87/  19.70 GFLOPS | Progress: (4/20) | 2.86 s
    [Task  5/25]  Current/Best:   15.38/  19.70 GFLOPS | Progress: (8/20) | 4.70 s
    [Task  5/25]  Current/Best:   20.58/  20.58 GFLOPS | Progress: (12/20) | 6.48 s
    [Task  5/25]  Current/Best:   18.06/  20.58 GFLOPS | Progress: (16/20) | 8.59 s
    [Task  5/25]  Current/Best:   13.01/  20.58 GFLOPS | Progress: (20/20) | 10.67 s Done.
-
    [Task  6/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  6/25]  Current/Best:   14.33/  14.33 GFLOPS | Progress: (4/20) | 5.51 s
    [Task  6/25]  Current/Best:    4.06/  16.30 GFLOPS | Progress: (8/20) | 9.09 s
    [Task  6/25]  Current/Best:   13.88/  18.00 GFLOPS | Progress: (12/20) | 11.51 s
    [Task  6/25]  Current/Best:    5.40/  18.00 GFLOPS | Progress: (16/20) | 18.65 s
    [Task  6/25]  Current/Best:   18.29/  18.29 GFLOPS | Progress: (20/20) | 20.96 s Done.
-
    [Task  7/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  7/25]  Current/Best:   15.20/  15.20 GFLOPS | Progress: (4/20) | 3.46 s
    [Task  7/25]  Current/Best:    6.23/  19.30 GFLOPS | Progress: (8/20) | 5.28 s
    [Task  7/25]  Current/Best:    5.76/  19.56 GFLOPS | Progress: (12/20) | 7.38 s
    [Task  7/25]  Current/Best:   12.47/  19.56 GFLOPS | Progress: (16/20) | 9.66 s
    [Task  7/25]  Current/Best:   12.35/  19.56 GFLOPS | Progress: (20/20) | 12.21 s Done.
-
    [Task  8/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  8/25]  Current/Best:   11.96/  17.28 GFLOPS | Progress: (4/20) | 4.86 s
    [Task  8/25]  Current/Best:   11.46/  17.28 GFLOPS | Progress: (8/20) | 12.13 s
    [Task  8/25]  Current/Best:    4.75/  17.28 GFLOPS | Progress: (12/20) | 18.56 s
    [Task  8/25]  Current/Best:   11.91/  17.37 GFLOPS | Progress: (16/20) | 21.37 s
    [Task  8/25]  Current/Best:    7.64/  17.37 GFLOPS | Progress: (20/20) | 25.35 s Done.
-
    [Task  9/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  9/25]  Current/Best:    1.89/  12.86 GFLOPS | Progress: (4/20) | 3.63 s
    [Task  9/25]  Current/Best:   11.89/  16.17 GFLOPS | Progress: (8/20) | 5.42 s
    [Task  9/25]  Current/Best:   12.02/  16.17 GFLOPS | Progress: (12/20) | 10.23 s
    [Task  9/25]  Current/Best:   15.31/  17.52 GFLOPS | Progress: (16/20) | 12.62 s
    [Task  9/25]  Current/Best:   20.09/  20.09 GFLOPS | Progress: (20/20) | 14.54 s Done.
-
    [Task 10/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 10/25]  Current/Best:   14.32/  14.85 GFLOPS | Progress: (4/20) | 3.45 s
    [Task 10/25]  Current/Best:   11.53/  18.58 GFLOPS | Progress: (8/20) | 4.76 s
    [Task 10/25]  Current/Best:    8.58/  18.58 GFLOPS | Progress: (12/20) | 6.32 s
    [Task 10/25]  Current/Best:   14.22/  18.58 GFLOPS | Progress: (16/20) | 8.33 s
    [Task 10/25]  Current/Best:   11.08/  18.58 GFLOPS | Progress: (20/20) | 9.96 s Done.
-
    [Task 11/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 11/25]  Current/Best:    3.12/  17.00 GFLOPS | Progress: (4/20) | 4.80 s
    [Task 11/25]  Current/Best:   16.82/  18.12 GFLOPS | Progress: (8/20) | 7.19 s
    [Task 11/25]  Current/Best:   12.34/  18.12 GFLOPS | Progress: (12/20) | 10.09 s
    [Task 11/25]  Current/Best:    1.57/  18.12 GFLOPS | Progress: (16/20) | 13.71 s
    [Task 11/25]  Current/Best:   20.54/  20.54 GFLOPS | Progress: (20/20) | 16.01 s Done.
-
    [Task 12/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 12/25]  Current/Best:    9.56/  15.70 GFLOPS | Progress: (4/20) | 10.39 s
    [Task 12/25]  Current/Best:   10.31/  18.76 GFLOPS | Progress: (8/20) | 12.35 s
    [Task 12/25]  Current/Best:   11.86/  18.76 GFLOPS | Progress: (12/20) | 17.78 s
    [Task 12/25]  Current/Best:    7.60/  18.76 GFLOPS | Progress: (16/20) | 24.01 s
    [Task 12/25]  Current/Best:   13.71/  18.76 GFLOPS | Progress: (20/20) | 26.60 s Done.
-
    [Task 13/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 13/25]  Current/Best:    5.12/  20.59 GFLOPS | Progress: (4/20) | 4.05 s
    [Task 13/25]  Current/Best:    6.12/  20.67 GFLOPS | Progress: (8/20) | 7.20 s
    [Task 13/25]  Current/Best:   12.08/  20.67 GFLOPS | Progress: (12/20) | 9.34 s
    [Task 13/25]  Current/Best:   14.34/  20.67 GFLOPS | Progress: (16/20) | 11.51 s
    [Task 13/25]  Current/Best:    9.25/  21.71 GFLOPS | Progress: (20/20) | 14.90 s Done.
-
    [Task 14/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 14/25]  Current/Best:    7.44/  11.29 GFLOPS | Progress: (4/20) | 6.25 s
    [Task 14/25]  Current/Best:   12.48/  12.48 GFLOPS | Progress: (8/20) | 8.62 s
    [Task 14/25]  Current/Best:   11.81/  13.03 GFLOPS | Progress: (12/20) | 13.01 s
    [Task 14/25]  Current/Best:   12.45/  13.03 GFLOPS | Progress: (16/20) | 18.03 s
    [Task 14/25]  Current/Best:   14.63/  14.63 GFLOPS | Progress: (20/20) | 23.32 s
    [Task 15/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 15/25]  Current/Best:   13.92/  16.77 GFLOPS | Progress: (4/20) | 3.48 s
    [Task 15/25]  Current/Best:   13.28/  19.09 GFLOPS | Progress: (8/20) | 5.91 s
    [Task 15/25]  Current/Best:   20.75/  20.75 GFLOPS | Progress: (12/20) | 7.20 s
    [Task 15/25]  Current/Best:   15.44/  20.75 GFLOPS | Progress: (16/20) | 9.49 s
    [Task 15/25]  Current/Best:   15.76/  20.75 GFLOPS | Progress: (20/20) 
 | 10.71 s
    [Task 16/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 16/25]  Current/Best:    5.97/  14.01 GFLOPS | Progress: (4/20) | 3.62 s Done.
+
    [Task  1/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  1/25]  Current/Best:   15.05/  23.80 GFLOPS | Progress: (4/20) | 6.92 s
    [Task  1/25]  Current/Best:   21.97/  23.80 GFLOPS | Progress: (8/20) | 10.22 s
    [Task  1/25]  Current/Best:   23.34/  23.80 GFLOPS | Progress: (12/20) | 16.23 s
    [Task  1/25]  Current/Best:   15.60/  23.80 GFLOPS | Progress: (16/20) | 18.93 s
    [Task  1/25]  Current/Best:   11.13/  23.80 GFLOPS | Progress: (20/20) | 20.94 s Done.
+
    [Task  2/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  2/25]  Current/Best:    7.90/  17.91 GFLOPS | Progress: (4/20) | 2.90 s
    [Task  2/25]  Current/Best:   18.50/  18.50 GFLOPS | Progress: (8/20) | 3.85 s
    [Task  2/25]  Current/Best:    3.32/  18.50 GFLOPS | Progress: (12/20) | 5.01 s
    [Task  2/25]  Current/Best:   14.53/  18.50 GFLOPS | Progress: (16/20) | 7.80 s
    [Task  2/25]  Current/Best:    8.66/  18.50 GFLOPS | Progress: (20/20) | 9.17 s Done.
+
    [Task  3/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  3/25]  Current/Best:   14.96/  14.96 GFLOPS | Progress: (4/20) | 3.55 s
    [Task  3/25]  Current/Best:   13.00/  14.96 GFLOPS | Progress: (8/20) | 5.83 s
    [Task  3/25]  Current/Best:   13.74/  17.98 GFLOPS | Progress: (12/20) | 8.25 s
    [Task  3/25]  Current/Best:    9.13/  23.69 GFLOPS | Progress: (16/20) | 10.97 s
    [Task  3/25]  Current/Best:   14.67/  23.69 GFLOPS | Progress: (20/20) | 12.98 s Done.
+
    [Task  4/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  4/25]  Current/Best:   12.29/  18.72 GFLOPS | Progress: (4/20) | 6.03 s
    [Task  4/25]  Current/Best:    3.38/  18.72 GFLOPS | Progress: (8/20) | 8.95 s
    [Task  4/25]  Current/Best:    7.92/  19.09 GFLOPS | Progress: (12/20) | 11.03 s
    [Task  4/25]  Current/Best:   19.22/  19.22 GFLOPS | Progress: (16/20) | 17.10 s
    [Task  4/25]  Current/Best:   17.34/  19.22 GFLOPS | Progress: (20/20) | 18.37 s Done.
+
    [Task  5/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  5/25]  Current/Best:    9.49/  19.34 GFLOPS | Progress: (4/20) | 3.50 s
    [Task  5/25]  Current/Best:   15.98/  19.34 GFLOPS | Progress: (8/20) | 5.22 s
    [Task  5/25]  Current/Best:   11.01/  20.06 GFLOPS | Progress: (12/20) | 8.19 s
    [Task  5/25]  Current/Best:   19.88/  20.06 GFLOPS | Progress: (16/20) | 10.39 s
    [Task  5/25]  Current/Best:   14.48/  20.06 GFLOPS | Progress: (20/20) | 12.38 s Done.
+
    [Task  6/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  6/25]  Current/Best:    7.16/  12.92 GFLOPS | Progress: (4/20) | 4.31 s
    [Task  6/25]  Current/Best:    7.94/  14.62 GFLOPS | Progress: (8/20) | 6.69 s
    [Task  6/25]  Current/Best:   14.69/  14.69 GFLOPS | Progress: (12/20) | 10.87 s
    [Task  6/25]  Current/Best:    8.36/  14.69 GFLOPS | Progress: (16/20) | 13.61 s
    [Task  6/25]  Current/Best:   13.55/  14.69 GFLOPS | Progress: (20/20) | 16.12 s Done.
+
    [Task  7/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  7/25]  Current/Best:    5.30/  19.41 GFLOPS | Progress: (4/20) | 3.88 s
    [Task  7/25]  Current/Best:   17.39/  19.41 GFLOPS | Progress: (8/20) | 5.92 s
    [Task  7/25]  Current/Best:   14.12/  19.41 GFLOPS | Progress: (12/20) | 9.25 s
    [Task  7/25]  Current/Best:   14.46/  20.93 GFLOPS | Progress: (16/20) | 12.96 s
    [Task  7/25]  Current/Best:   11.59/  20.93 GFLOPS | Progress: (20/20) | 15.61 s Done.
+
    [Task  8/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  8/25]  Current/Best:    9.95/  17.55 GFLOPS | Progress: (4/20) | 4.78 s
    [Task  8/25]  Current/Best:    9.71/  17.55 GFLOPS | Progress: (8/20) | 6.96 s
    [Task  8/25]  Current/Best:   11.13/  18.83 GFLOPS | Progress: (12/20) | 9.31 s
    [Task  8/25]  Current/Best:   16.02/  18.83 GFLOPS | Progress: (16/20) | 11.50 s
    [Task  8/25]  Current/Best:   11.55/  18.83 GFLOPS | Progress: (20/20) | 13.76 s Done.
+
    [Task  9/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  9/25]  Current/Best:   12.43/  18.23 GFLOPS | Progress: (4/20) | 2.95 s
    [Task  9/25]  Current/Best:   12.34/  18.23 GFLOPS | Progress: (8/20) | 4.85 s
    [Task  9/25]  Current/Best:   17.50/  18.23 GFLOPS | Progress: (12/20) | 8.82 s
    [Task  9/25]  Current/Best:    4.53/  18.23 GFLOPS | Progress: (16/20) | 10.48 s
    [Task  9/25]  Current/Best:   17.16/  18.23 GFLOPS | Progress: (20/20) | 12.97 s Done.
+
    [Task 10/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 10/25]  Current/Best:   13.34/  13.34 GFLOPS | Progress: (4/20) | 3.18 s
    [Task 10/25]  Current/Best:   18.36/  18.36 GFLOPS | Progress: (8/20) | 4.76 s
    [Task 10/25]  Current/Best:   10.32/  19.02 GFLOPS | Progress: (12/20) | 6.81 s
    [Task 10/25]  Current/Best:    4.68/  19.02 GFLOPS | Progress: (16/20) | 8.48 s
    [Task 10/25]  Current/Best:   10.74/  19.02 GFLOPS | Progress: (20/20) | 11.13 s Done.
+
    [Task 11/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 11/25]  Current/Best:   18.27/  18.27 GFLOPS | Progress: (4/20) | 3.58 s
    [Task 11/25]  Current/Best:    7.78/  18.27 GFLOPS | Progress: (8/20) | 6.25 s
    [Task 11/25]  Current/Best:   11.59/  18.98 GFLOPS | Progress: (12/20) | 9.74 s
    [Task 11/25]  Current/Best:   19.05/  19.05 GFLOPS | Progress: (16/20) | 11.89 s
    [Task 11/25]  Current/Best:   10.89/  19.05 GFLOPS | Progress: (20/20) | 14.33 s Done.
+
    [Task 12/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 12/25]  Current/Best:   14.58/  14.58 GFLOPS | Progress: (4/20) | 3.72 s
    [Task 12/25]  Current/Best:   18.08/  18.08 GFLOPS | Progress: (8/20) | 6.95 s
    [Task 12/25]  Current/Best:    8.90/  18.08 GFLOPS | Progress: (12/20) | 10.81 s
    [Task 12/25]  Current/Best:    8.18/  22.41 GFLOPS | Progress: (16/20) | 16.04 s
    [Task 12/25]  Current/Best:   11.51/  22.41 GFLOPS | Progress: (20/20) | 18.62 s Done.
+
    [Task 13/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 13/25]  Current/Best:    6.98/  12.20 GFLOPS | Progress: (4/20) | 4.76 s
    [Task 13/25]  Current/Best:   17.33/  17.33 GFLOPS | Progress: (8/20) | 8.29 s
    [Task 13/25]  Current/Best:   16.65/  17.33 GFLOPS | Progress: (12/20) | 11.16 s
    [Task 13/25]  Current/Best:   11.19/  21.89 GFLOPS | Progress: (16/20) | 14.44 s
    [Task 13/25]  Current/Best:   19.44/  21.89 GFLOPS | Progress: (20/20) | 17.26 s Done.
+
    [Task 14/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 14/25]  Current/Best:   19.91/  19.91 GFLOPS | Progress: (4/20) | 4.73 s
    [Task 14/25]  Current/Best:    4.86/  19.91 GFLOPS | Progress: (8/20) | 7.75 s
    [Task 14/25]  Current/Best:   11.85/  19.91 GFLOPS | Progress: (12/20) | 10.42 s
    [Task 14/25]  Current/Best:   12.03/  21.32 GFLOPS | Progress: (16/20) | 14.02 s
    [Task 14/25]  Current/Best:   11.51/  21.32 GFLOPS | Progress: (20/20) | 16.04 s
    [Task 15/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+
    [Task 15/25]  Current/Best:    3.20/  17.75 GFLOPS | Progress: (4/20) | 3.77 s
    [Task 15/25]  Current/Best:   18.27/  21.63 GFLOPS | Progress: (8/20) | 6.45 s
    [Task 15/25]  Current/Best:   11.55/  21.63 GFLOPS | Progress: (12/20) | 10.22 s
    [Task 15/25]  Current/Best:   19.05/  21.63 GFLOPS | Progress: (16/20) | 12.13 s
    [Task 15/25]  Current/Best:   12.32/  21.63 GFLOPS | Progress: (20/20) | 14.32 s
    [Task 16/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 16/25]  Current/Best:   15.80/  17.62 GFLOPS | Progress: (4/20) | 3.71 s
    [Task 16/25]  Current/Best:   20.37/  20.37 GFLOPS | Progress: (8/20) | 5.51 s
    [Task 16/25]  Current/Best:   22.67/  22.67 GFLOPS | Progress: (12/20) | 7.03 s
    [Task 16/25]  Current/Best:   16.72/  22.67 GFLOPS | Progress: (16/20) | 9.52 s
    [Task 16/25]  Current/Best:   17.86/  22.67 GFLOPS | Progress: (20/20) | 11.82 s Done.
+
    [Task 17/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 17/25]  Current/Best:    6.18/  15.78 GFLOPS | Progress: (4/20) | 4.35 s
    [Task 17/25]  Current/Best:    9.64/  15.78 GFLOPS | Progress: (8/20) | 7.96 s
    [Task 17/25]  Current/Best:    6.03/  19.53 GFLOPS | Progress: (12/20) | 11.37 s
    [Task 17/25]  Current/Best:   21.62/  21.62 GFLOPS | Progress: (16/20) | 14.65 s
    [Task 17/25]  Current/Best:   13.06/  21.62 GFLOPS | Progress: (20/20) | 16.92 s Done.
+
    [Task 18/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 18/25]  Current/Best:    2.99/  16.50 GFLOPS | Progress: (4/20) | 3.83 s
    [Task 18/25]  Current/Best:   10.49/  20.92 GFLOPS | Progress: (8/20) | 10.24 s
    [Task 18/25]  Current/Best:    9.57/  20.92 GFLOPS | Progress: (12/20) | 17.95 s
    [Task 18/25]  Current/Best:   14.32/  20.92 GFLOPS | Progress: (16/20) | 19.68 s
    [Task 18/25]  Current/Best:   15.02/  20.92 GFLOPS | Progress: (20/20) | 22.34 s Done.
+
    [Task 19/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 19/25]  Current/Best:    7.39/  17.00 GFLOPS | Progress: (4/20) | 4.71 s
    [Task 19/25]  Current/Best:   18.49/  18.49 GFLOPS | Progress: (8/20) | 8.50 s
    [Task 19/25]  Current/Best:    9.84/  18.49 GFLOPS | Progress: (12/20) | 12.85 s
    [Task 19/25]  Current/Best:   15.24/  18.49 GFLOPS | Progress: (16/20) | 17.12 s
    [Task 19/25]  Current/Best:    7.83/  18.49 GFLOPS | Progress: (20/20) | 21.34 s Done.
+
    [Task 20/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 20/25]  Current/Best:   11.56/  17.55 GFLOPS | Progress: (4/20) | 4.13 s
    [Task 20/25]  Current/Best:    2.71/  19.22 GFLOPS | Progress: (8/20) | 6.47 s
    [Task 20/25]  Current/Best:   15.37/  19.22 GFLOPS | Progress: (12/20) | 9.01 s
    [Task 20/25]  Current/Best:   13.83/  20.14 GFLOPS | Progress: (16/20) | 11.23 s
    [Task 20/25]  Current/Best:   12.42/  20.14 GFLOPS | Progress: (20/20) | 14.07 s
    [Task 21/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 21/25]  Current/Best:    7.06/  15.55 GFLOPS | Progress: (4/20) | 2.98 s
    [Task 21/25]  Current/Best:    9.11/  15.55 GFLOPS | Progress: (8/20) | 6.48 s
    [Task 21/25]  Current/Best:   20.61/  20.61 GFLOPS | Progress: (12/20) | 9.92 s
    [Task 21/25]  Current/Best:   18.12/  21.48 GFLOPS | Progress: (16/20) | 11.09 s Done.
      Done.
-
    [Task 16/25]  Current/Best:   15.54/  20.34 GFLOPS | Progress: (8/20) | 6.49 s
    [Task 16/25]  Current/Best:   19.30/  20.34 GFLOPS | Progress: (12/20) | 8.06 s
    [Task 16/25]  Current/Best:   14.24/  20.34 GFLOPS | Progress: (16/20) | 10.58 s
    [Task 16/25]  Current/Best:    4.57/  20.34 GFLOPS | Progress: (20/20) | 12.09 s Done.
-
    [Task 17/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 17/25]  Current/Best:   15.16/  18.26 GFLOPS | Progress: (4/20) | 3.45 s
    [Task 17/25]  Current/Best:   17.30/  20.30 GFLOPS | Progress: (8/20) | 5.45 s
    [Task 17/25]  Current/Best:   18.28/  20.30 GFLOPS | Progress: (12/20) | 7.76 s
    [Task 17/25]  Current/Best:   22.24/  22.24 GFLOPS | Progress: (16/20) | 9.28 s
    [Task 17/25]  Current/Best:   15.56/  22.24 GFLOPS | Progress: (20/20) | 11.75 s Done.
-
    [Task 18/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 18/25]  Current/Best:   15.15/  15.15 GFLOPS | Progress: (4/20) | 6.83 s
    [Task 18/25]  Current/Best:    9.94/  17.98 GFLOPS | Progress: (8/20) | 11.68 s
    [Task 18/25]  Current/Best:   10.07/  17.98 GFLOPS | Progress: (12/20) | 15.33 s
    [Task 18/25]  Current/Best:   11.48/  17.98 GFLOPS | Progress: (16/20) | 18.40 s
    [Task 18/25]  Current/Best:    5.07/  19.22 GFLOPS | Progress: (20/20) | 20.34 s Done.
-
    [Task 19/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 19/25]  Current/Best:    8.42/  21.05 GFLOPS | Progress: (4/20) | 6.45 s
    [Task 19/25]  Current/Best:   19.58/  21.05 GFLOPS | Progress: (8/20) | 9.99 s
    [Task 19/25]  Current/Best:   18.15/  21.05 GFLOPS | Progress: (12/20) | 12.09 s
    [Task 19/25]  Current/Best:   16.87/  21.05 GFLOPS | Progress: (16/20) | 15.33 s
    [Task 19/25]  Current/Best:    9.35/  21.05 GFLOPS | Progress: (20/20) | 19.20 s Done.
-
    [Task 20/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 20/25]  Current/Best:    9.72/  17.55 GFLOPS | Progress: (4/20) | 3.03 s
    [Task 20/25]  Current/Best:   13.98/  17.55 GFLOPS | Progress: (8/20) | 6.28 s
    [Task 20/25]  Current/Best:   10.40/  17.55 GFLOPS | Progress: (12/20) | 8.90 s
    [Task 20/25]  Current/Best:   17.84/  20.50 GFLOPS | Progress: (16/20) | 11.69 s
    [Task 20/25]  Current/Best:   15.56/  20.50 GFLOPS | Progress: (20/20) | 15.71 s
    [Task 21/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
    [Task 21/25]  Current/Best:    2.52/  12.52 GFLOPS | Progress: (4/20) | 3.34 s
    [Task 21/25]  Current/Best:   17.76/  17.76 GFLOPS | Progress: (8/20) | 7.84 s
    [Task 21/25]  Current/Best:    5.25/  17.76 GFLOPS | Progress: (12/20) | 9.86 s
    [Task 21/25]  Current/Best:    2.68/  21.66 GFLOPS | Progress: (16/20) | 12.26 s
    [Task 21/25]  Current/Best:   13.50/  21.66 GFLOPS | Progress: (20/20) | 13.87 s
    [Task 22/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 22/25]  Current/Best:    5.19/  12.99 GFLOPS | Progress: (4/20) | 3.85 s
    [Task 22/25]  Current/Best:    7.63/  16.16 GFLOPS | Progress: (8/20) | 5.67 s
    [Task 22/25]  Current/Best:    9.04/  18.37 GFLOPS | Progress: (12/20) | 8.13 s
    [Task 22/25]  Current/Best:   10.87/  18.37 GFLOPS | Progress: (16/20) | 11.44 s
    [Task 22/25]  Current/Best:    7.64/  18.37 GFLOPS | Progress: (20/20) | 13.54 s Done.
-
    [Task 23/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 23/25]  Current/Best:    1.55/  20.48 GFLOPS | Progress: (4/20) | 5.00 s
    [Task 23/25]  Current/Best:    5.36/  23.62 GFLOPS | Progress: (8/20) | 9.23 s
    [Task 23/25]  Current/Best:   11.24/  23.62 GFLOPS | Progress: (12/20) | 12.45 s
    [Task 23/25]  Current/Best:    8.43/  23.62 GFLOPS | Progress: (16/20) | 15.74 s
    [Task 23/25]  Current/Best:   11.42/  23.62 GFLOPS | Progress: (20/20) | 19.04 s Done.
-
    [Task 24/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 24/25]  Current/Best:    5.51/   5.51 GFLOPS | Progress: (4/20) | 4.97 s
    [Task 24/25]  Current/Best:    3.99/   5.51 GFLOPS | Progress: (8/20) | 15.70 s
    [Task 24/25]  Current/Best:    3.49/   8.12 GFLOPS | Progress: (12/20) | 26.45 s
    [Task 24/25]  Current/Best:    6.96/   8.12 GFLOPS | Progress: (16/20) | 37.16 s
    [Task 24/25]  Current/Best:    1.07/   8.12 GFLOPS | Progress: (20/20) | 44.66 s
    [Task 25/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
    [Task 25/25]  Current/Best:    3.52/   5.27 GFLOPS | Progress: (4/20) | 2.61 s
    [Task 25/25]  Current/Best:    8.89/   8.89 GFLOPS | Progress: (8/20) | 13.29 s
    [Task 25/25]  Current/Best:    3.63/   8.92 GFLOPS | Progress: (12/20) | 24.01 s
    [Task 25/25]  Current/Best:    1.51/   8.92 GFLOPS | Progress: (16/20) | 26.99 s
    [Task 25/25]  Current/Best:    5.81/   8.92 GFLOPS | Progress: (20/20) | 28.02 s
+
    [Task 21/25]  Current/Best:   10.95/  21.69 GFLOPS | Progress: (20/20) | 13.57 s Done.
+
    [Task 22/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 22/25]  Current/Best:   12.03/  17.70 GFLOPS | Progress: (4/20) | 2.87 s
    [Task 22/25]  Current/Best:    9.89/  17.70 GFLOPS | Progress: (8/20) | 4.33 s
    [Task 22/25]  Current/Best:   18.35/  18.35 GFLOPS | Progress: (12/20) | 5.95 s
    [Task 22/25]  Current/Best:    6.14/  18.35 GFLOPS | Progress: (16/20) | 8.24 s
    [Task 22/25]  Current/Best:   12.32/  18.35 GFLOPS | Progress: (20/20) | 10.35 s Done.
+
    [Task 23/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 23/25]  Current/Best:   21.15/  21.15 GFLOPS | Progress: (4/20) | 3.56 s
    [Task 23/25]  Current/Best:    3.08/  21.15 GFLOPS | Progress: (8/20) | 7.43 s
    [Task 23/25]  Current/Best:   12.91/  21.15 GFLOPS | Progress: (12/20) | 9.98 s
    [Task 23/25]  Current/Best:   14.52/  21.45 GFLOPS | Progress: (16/20) | 12.58 s
    [Task 23/25]  Current/Best:    9.85/  21.45 GFLOPS | Progress: (20/20) | 15.09 s Done.
+
    [Task 24/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 24/25]  Current/Best:    5.72/   9.83 GFLOPS | Progress: (4/20) | 4.45 s
    [Task 24/25]  Current/Best:    3.19/   9.83 GFLOPS | Progress: (8/20) | 8.93 s
    [Task 24/25]  Current/Best:    3.25/   9.83 GFLOPS | Progress: (12/20) | 10.79 s
    [Task 24/25]  Current/Best:    1.20/   9.83 GFLOPS | Progress: (16/20) | 21.51 s
    [Task 24/25]  Current/Best:    2.94/   9.83 GFLOPS | Progress: (20/20) | 32.38 s
    [Task 25/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 25/25]  Current/Best:    1.55/   3.42 GFLOPS | Progress: (4/20) | 12.33 s
    [Task 25/25]  Current/Best:    5.52/   5.52 GFLOPS | Progress: (8/20) | 17.66 s
    [Task 25/25]  Current/Best:    7.17/   7.76 GFLOPS | Progress: (12/20) | 21.97 s
    [Task 25/25]  Current/Best:    2.73/   7.76 GFLOPS | Progress: (16/20) | 32.72 s Done.
+
    [Task 25/25]  Current/Best:    4.83/   7.76 GFLOPS | Progress: (20/20) | 36.69 s Done.
+
 
 
 
@@ -731,8 +733,8 @@ improvement in comparing the optimized model to the unoptimized model.
 
  .. code-block:: none
 
-    optimized: {'mean': 410.15801464000106, 'median': 408.6431573000027, 'std': 3.983016703746832}
-    unoptimized: {'mean': 516.8028887700029, 'median': 516.7933151499938, 'std': 1.8442027952596247}
+    optimized: {'mean': 429.7182228099973, 'median': 431.4217151000321, 'std': 4.16696520745253}
+    unoptimized: {'mean': 521.135307909999, 'median': 521.1086226000134, 'std': 2.045413370611934}
 
 
 
@@ -755,7 +757,7 @@ profiling/benchmarking.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 10 minutes  53.937 seconds)
+   **Total running time of the script:** ( 10 minutes  33.992 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 8c7a73aa71..3997ce2d64 100644
--- a/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
+++ b/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
@@ -270,7 +270,7 @@ device and returns the measured cost. Network overhead is excluded.
 
  .. code-block:: none
 
-    1.227e-07 secs/op
+    1.264e-07 secs/op
 
 
 
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index 5d7dfc97f2..d5dc89805f 100644
--- a/docs/_sources/tutorial/intro_topi.rst.txt
+++ b/docs/_sources/tutorial/intro_topi.rst.txt
@@ -260,7 +260,7 @@ As you can see, scheduled stages of computation have been accumulated and we can
 
  .. code-block:: none
 
-    [stage(a, placeholder(a, 0x22ef0670)), stage(b, placeholder(b, 0xe875ed0)), stage(T_add, compute(T_add, body=[(a[ax0, ax1, ax2] + b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min=0, ext=10))], reduce_axis=[], tag=broadcast, attrs={})), stage(T_multiply, compute(T_multiply, body=[(a[ax0, ax1, ax2]*b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min [...]
+    [stage(a, placeholder(a, 0x20961e90)), stage(b, placeholder(b, 0x21e64170)), stage(T_add, compute(T_add, body=[(a[ax0, ax1, ax2] + b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min=0, ext=10))], reduce_axis=[], tag=broadcast, attrs={})), stage(T_multiply, compute(T_multiply, body=[(a[ax0, ax1, ax2]*b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(mi [...]
 
 
 
diff --git a/docs/_sources/tutorial/sg_execution_times.rst.txt b/docs/_sources/tutorial/sg_execution_times.rst.txt
index a959ae05b7..8e97fff97f 100644
--- a/docs/_sources/tutorial/sg_execution_times.rst.txt
+++ b/docs/_sources/tutorial/sg_execution_times.rst.txt
@@ -5,26 +5,26 @@
 
 Computation times
 =================
-**14:03.663** total execution time for **tutorial** files:
+**14:00.891** total execution time for **tutorial** files:
 
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``)                 | 10:53.937 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``)                 | 10:33.992 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:16.007 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:21.583 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``)     | 01:01.084 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``)     | 01:00.939 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``)                 | 00:35.950 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``)                 | 00:36.630 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``)               | 00:15.132 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``)               | 00:26.073 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``)                               | 00:00.771 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``)                               | 00:00.799 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``)       | 00:00.628 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``)       | 00:00.668 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.145 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.194 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``)                           | 00:00.005 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``)                           | 00:00.007 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_tutorial_uma.py` (``uma.py``)                                             | 00:00.002 | 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 bcf3e2af07..dd0da46ae0 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -294,8 +294,8 @@ helper function to run a profile of the TVM generated code.
 
  .. code-block:: none
 
-    Numpy running time: 0.000007
-    naive: 0.000007
+    Numpy running time: 0.000008
+    naive: 0.000010
 
 
 
@@ -393,7 +393,7 @@ compile and run this new schedule with the parallel operation applied:
 
  .. code-block:: none
 
-    parallel: 0.000007
+    parallel: 0.000009
 
 
 
@@ -499,10 +499,10 @@ We can now compare the different schedules
  .. code-block:: none
 
                 Operator                  Timing             Performance
-                   numpy    7.26831000065431e-06                     1.0
-                   naive    6.705600000000001e-06     0.9225803521583898
-                parallel              6.9417e-06      0.9550638318089201
-                  vector             2.45513e-05      3.3778553746042523
+                   numpy    7.852259996070643e-06                    1.0
+                   naive              9.7594e-06       1.242877847254639
+                parallel              9.3415e-06      1.1896575004743335
+                  vector    2.4529200000000002e-05    3.1238395076417085
 
 
 
@@ -923,7 +923,7 @@ matrix multiplication.
 
  .. code-block:: none
 
-    Numpy running time: 0.018680
+    Numpy running time: 0.019926
 
 
 
@@ -981,7 +981,7 @@ optimizations.
 
  .. code-block:: none
 
-    none: 3.420528
+    none: 3.330825
 
 
 
@@ -1083,7 +1083,7 @@ schedule.
 
  .. code-block:: none
 
-    blocking: 0.298977
+    blocking: 0.330132
 
 
 
@@ -1178,7 +1178,7 @@ already cache friendly from our previous optimizations.
 
  .. code-block:: none
 
-    vectorization: 0.333744
+    vectorization: 0.363457
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1251,7 +1251,7 @@ more cache friendly.
 
  .. code-block:: none
 
-    loop permutation: 0.117542
+    loop permutation: 0.133229
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1349,7 +1349,7 @@ optimized schedule.
 
  .. code-block:: none
 
-    array packing: 0.108424
+    array packing: 0.107371
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1441,7 +1441,7 @@ to `C` when all the block results are ready.
 
  .. code-block:: none
 
-    block caching: 0.110747
+    block caching: 0.111066
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1526,7 +1526,7 @@ of thread-level parallelization.
 
  .. code-block:: none
 
-    parallelization: 0.146593
+    parallelization: 0.146867
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1606,13 +1606,13 @@ working, we can compare the results.
  .. code-block:: none
 
                 Operator                  Timing             Performance
-                    none            3.4205275692                     1.0
-                blocking     0.29897708349999996     0.08740671649371484
-           vectorization             0.333743835     0.09757086538497238
-        loop permutation            0.1175421273    0.034363742119316115
-           array packing     0.10842431959999999     0.03169812767372563
-           block caching     0.11074702389999999     0.03237717622778925
-         parallelization     0.14659330210000002    0.042856927516092366
+                    none            3.3308246923                     1.0
+                blocking             0.330131613     0.09911407639170515
+           vectorization            0.3634572827     0.10911930716144223
+        loop permutation            0.1332293653     0.03999891246392872
+           array packing            0.1073706838     0.03223546530329654
+           block caching     0.11106572819999998    0.033344813510226175
+         parallelization            0.1468670843     0.04409330957570913
 
 
 
@@ -1654,7 +1654,7 @@ the computation for specific platforms.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  1.084 seconds)
+   **Total running time of the script:** ( 1 minutes  0.939 seconds)
 
 
 .. _sphx_glr_download_tutorial_tensor_expr_get_started.py:
diff --git a/docs/commit_hash b/docs/commit_hash
index c8590f0509..78287c1c83 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-44ed06ac9f019f9f06608504c3382d0905b6d5a2
+52739ef8cd4499b555453d7159991316afc44301
diff --git a/docs/how_to/compile_models/from_darknet.html b/docs/how_to/compile_models/from_darknet.html
index b7e046ed8b..1e3e30ba57 100644
--- a/docs/how_to/compile_models/from_darknet.html
+++ b/docs/how_to/compile_models/from_darknet.html
@@ -585,7 +585,7 @@ class:[&#39;truck 0.9266&#39;] left:471 top:83 right:689 bottom:169
 class:[&#39;bicycle 0.9984&#39;] 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  13.405 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  16.228 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 997de5d4b5..6852fe7b67 100644
--- a/docs/how_to/compile_models/from_keras.html
+++ b/docs/how_to/compile_models/from_keras.html
@@ -506,7 +506,7 @@ pip install -U tensorflow --user
 <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 964ms/step
+1/1 [==============================] - 1s 955ms/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 3076dc5d23..b2eade0ba9 100644
--- a/docs/how_to/compile_models/from_mxnet.html
+++ b/docs/how_to/compile_models/from_mxnet.html
@@ -440,7 +440,7 @@ to download the full example code</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;x&quot;</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.zip6703bb06-6b2d-4042-9067-f67e09bca4e5 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.zip2b781813-d81c-436a-a775-b37a76c5a5e2 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 10c3f37b19..0b7745e302 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -448,14 +448,15 @@ 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: &quot;https://oneflow-public.oss-cn-beijing.aliyuncs.com/model_zoo/flowvision/classification/ResNet/resnet18.zip&quot; to /workspace/.oneflow/flowvision_cache/resnet18.zip
 
   0%|          | 0.00/41.5M [00:00&lt;?, ?B/s]
- 15%|#5        | 6.33M/41.5M [00:00&lt;00:00, 37.1MB/s]
- 24%|##3       | 9.87M/41.5M [00:00&lt;00:01, 29.5MB/s]
- 39%|###8      | 16.0M/41.5M [00:00&lt;00:00, 33.4MB/s]
- 54%|#####3    | 22.3M/41.5M [00:00&lt;00:00, 42.2MB/s]
- 73%|#######3  | 30.3M/41.5M [00:00&lt;00:00, 54.0MB/s]
- 87%|########6 | 36.0M/41.5M [00:00&lt;00:00, 45.8MB/s]
- 98%|#########8| 40.8M/41.5M [00:01&lt;00:00, 43.9MB/s]
-100%|##########| 41.5M/41.5M [00:01&lt;00:00, 42.9MB/s]
+ 15%|#5        | 6.33M/41.5M [00:00&lt;00:01, 31.4MB/s]
+ 27%|##6       | 11.0M/41.5M [00:00&lt;00:00, 38.6MB/s]
+ 39%|###8      | 16.0M/41.5M [00:00&lt;00:00, 33.8MB/s]
+ 54%|#####3    | 22.3M/41.5M [00:00&lt;00:00, 33.7MB/s]
+ 62%|######1   | 25.7M/41.5M [00:00&lt;00:00, 31.0MB/s]
+ 77%|#######7  | 32.0M/41.5M [00:00&lt;00:00, 35.2MB/s]
+ 85%|########5 | 35.4M/41.5M [00:01&lt;00:00, 33.5MB/s]
+ 93%|#########3| 38.6M/41.5M [00:01&lt;00:00, 32.8MB/s]
+100%|##########| 41.5M/41.5M [00:01&lt;00:00, 34.3MB/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 3c70333604..29671fb9f3 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -431,14 +431,12 @@ be unstable.</p>
 Downloading: &quot;https://download.pytorch.org/models/resnet18-f37072fd.pth&quot; to /workspace/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
 
   0%|          | 0.00/44.7M [00:00&lt;?, ?B/s]
-  6%|5         | 2.61M/44.7M [00:00&lt;00:01, 22.4MB/s]
- 18%|#7        | 7.99M/44.7M [00:00&lt;00:01, 33.0MB/s]
- 32%|###2      | 14.3M/44.7M [00:00&lt;00:01, 31.7MB/s]
- 39%|###8      | 17.3M/44.7M [00:00&lt;00:00, 29.7MB/s]
- 55%|#####4    | 24.5M/44.7M [00:00&lt;00:00, 41.9MB/s]
- 72%|#######1  | 32.0M/44.7M [00:00&lt;00:00, 43.9MB/s]
- 90%|########9 | 40.0M/44.7M [00:01&lt;00:00, 42.2MB/s]
-100%|##########| 44.7M/44.7M [00:01&lt;00:00, 42.3MB/s]
+ 18%|#7        | 7.99M/44.7M [00:00&lt;00:00, 68.8MB/s]
+ 36%|###5      | 16.0M/44.7M [00:00&lt;00:00, 55.9MB/s]
+ 54%|#####3    | 24.0M/44.7M [00:00&lt;00:00, 62.2MB/s]
+ 72%|#######1  | 32.0M/44.7M [00:00&lt;00:00, 66.3MB/s]
+ 90%|########9 | 40.1M/44.7M [00:00&lt;00:00, 70.1MB/s]
+100%|##########| 44.7M/44.7M [00:00&lt;00:00, 71.9MB/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 89f5e05ea6..3977441680 100644
--- a/docs/how_to/compile_models/from_tensorflow.html
+++ b/docs/how_to/compile_models/from_tensorflow.html
@@ -645,7 +645,7 @@ 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  11.788 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  13.077 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 f7ba3d0669..77e074abe4 100644
--- a/docs/how_to/compile_models/sg_execution_times.html
+++ b/docs/how_to/compile_models/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-compile-models-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>05:52.981</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>05:58.074</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_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>01:13.405</p></td>
+<td><p>01:16.228</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><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>01:11.788</p></td>
+<td><p>01:13.077</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>00:46.658</p></td>
+<td><p>00:47.284</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>00:33.757</p></td>
+<td><p>00:33.815</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="from_mxnet.html#sphx-glr-how-to-compile-models-from-mxnet-py"><span class="std std-ref">Compile MXNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_mxnet.py</span></code>)</p></td>
-<td><p>00:30.473</p></td>
+<td><p>00:29.966</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="from_coreml.html#sphx-glr-how-to-compile-models-from-coreml-py"><span class="std std-ref">Compile CoreML Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_coreml.py</span></code>)</p></td>
-<td><p>00:26.763</p></td>
+<td><p>00:27.561</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="from_tflite.html#sphx-glr-how-to-compile-models-from-tflite-py"><span class="std std-ref">Compile TFLite Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tflite.py</span></code>)</p></td>
-<td><p>00:25.790</p></td>
+<td><p>00:26.150</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="from_pytorch.html#sphx-glr-how-to-compile-models-from-pytorch-py"><span class="std std-ref">Compile PyTorch Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_pytorch.py</span></code>)</p></td>
-<td><p>00:23.270</p></td>
+<td><p>00:23.095</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>00:18.636</p></td>
+<td><p>00:18.467</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>00:02.439</p></td>
+<td><p>00:02.433</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 </tbody>
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 b7f68b86fa..b823d3f3e9 100644
--- a/docs/how_to/deploy_models/deploy_model_on_android.html
+++ b/docs/how_to/deploy_models/deploy_model_on_android.html
@@ -661,7 +661,7 @@ to the remote android device.</p>
 Evaluate inference time cost...
 Execution time summary:
  mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)
-  16.2805      16.3876      16.6878      15.6962       0.3367
+  16.1529      16.1286      16.3120      16.0284       0.0906
 </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 19a38bad0c..54bddae2f0 100644
--- a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
+++ b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
@@ -453,32 +453,32 @@ be unstable.</p>
 Downloading: &quot;https://download.pytorch.org/models/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth&quot; to /workspace/.cache/torch/hub/checkpoints/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth
 
   0%|          | 0.00/170M [00:00&lt;?, ?B/s]
-  5%|4         | 7.99M/170M [00:00&lt;00:02, 61.7MB/s]
-  8%|8         | 14.3M/170M [00:00&lt;00:04, 36.6MB/s]
- 11%|#         | 18.2M/170M [00:00&lt;00:04, 37.0MB/s]
- 13%|#3        | 22.3M/170M [00:00&lt;00:05, 29.5MB/s]
- 15%|#4        | 25.4M/170M [00:00&lt;00:05, 29.7MB/s]
- 21%|##        | 35.3M/170M [00:00&lt;00:02, 48.6MB/s]
- 24%|##3       | 40.7M/170M [00:01&lt;00:02, 46.7MB/s]
- 28%|##8       | 48.0M/170M [00:01&lt;00:02, 48.3MB/s]
- 33%|###2      | 56.0M/170M [00:01&lt;00:02, 52.5MB/s]
- 38%|###7      | 64.0M/170M [00:01&lt;00:01, 58.0MB/s]
- 42%|####2     | 72.0M/170M [00:01&lt;00:01, 53.5MB/s]
- 47%|####7     | 80.0M/170M [00:01&lt;00:01, 53.3MB/s]
- 51%|#####     | 86.3M/170M [00:01&lt;00:01, 56.0MB/s]
- 54%|#####4    | 91.9M/170M [00:01&lt;00:01, 56.0MB/s]
- 58%|#####8    | 98.6M/170M [00:02&lt;00:01, 59.6MB/s]
- 62%|######1   | 104M/170M [00:02&lt;00:01, 54.0MB/s]
- 68%|######7   | 115M/170M [00:02&lt;00:00, 68.1MB/s]
- 72%|#######1  | 122M/170M [00:02&lt;00:00, 66.7MB/s]
- 76%|#######5  | 128M/170M [00:02&lt;00:00, 64.8MB/s]
- 80%|########  | 136M/170M [00:02&lt;00:00, 56.4MB/s]
- 83%|########3 | 142M/170M [00:02&lt;00:00, 51.1MB/s]
- 87%|########7 | 148M/170M [00:02&lt;00:00, 54.5MB/s]
- 90%|######### | 153M/170M [00:03&lt;00:00, 42.6MB/s]
- 94%|#########4| 160M/170M [00:03&lt;00:00, 45.1MB/s]
- 99%|#########8| 168M/170M [00:03&lt;00:00, 53.5MB/s]
-100%|##########| 170M/170M [00:03&lt;00:00, 51.6MB/s]
+  5%|4         | 7.99M/170M [00:00&lt;00:04, 39.9MB/s]
+  8%|8         | 14.3M/170M [00:00&lt;00:03, 40.8MB/s]
+ 11%|#         | 18.2M/170M [00:00&lt;00:04, 36.0MB/s]
+ 14%|#4        | 24.0M/170M [00:00&lt;00:04, 36.7MB/s]
+ 19%|#8        | 32.0M/170M [00:00&lt;00:03, 38.2MB/s]
+ 24%|##3       | 40.0M/170M [00:01&lt;00:03, 43.5MB/s]
+ 28%|##8       | 48.0M/170M [00:01&lt;00:02, 43.9MB/s]
+ 33%|###2      | 56.0M/170M [00:01&lt;00:02, 48.5MB/s]
+ 37%|###6      | 62.3M/170M [00:01&lt;00:02, 52.2MB/s]
+ 40%|###9      | 67.5M/170M [00:01&lt;00:02, 47.1MB/s]
+ 42%|####2     | 72.2M/170M [00:01&lt;00:02, 41.2MB/s]
+ 47%|####6     | 79.7M/170M [00:01&lt;00:01, 49.6MB/s]
+ 50%|####9     | 84.8M/170M [00:02&lt;00:01, 45.0MB/s]
+ 53%|#####2    | 89.5M/170M [00:02&lt;00:02, 33.8MB/s]
+ 56%|#####6    | 95.5M/170M [00:02&lt;00:01, 39.6MB/s]
+ 59%|#####8    | 99.9M/170M [00:02&lt;00:02, 35.7MB/s]
+ 61%|######1   | 104M/170M [00:02&lt;00:02, 31.7MB/s]
+ 66%|######5   | 112M/170M [00:02&lt;00:01, 35.3MB/s]
+ 71%|#######   | 120M/170M [00:03&lt;00:01, 43.0MB/s]
+ 75%|#######5  | 128M/170M [00:03&lt;00:00, 44.1MB/s]
+ 80%|########  | 136M/170M [00:03&lt;00:00, 46.6MB/s]
+ 85%|########4 | 144M/170M [00:03&lt;00:00, 45.1MB/s]
+ 89%|########9 | 152M/170M [00:03&lt;00:00, 45.9MB/s]
+ 93%|#########3| 158M/170M [00:03&lt;00:00, 45.5MB/s]
+ 96%|#########5| 163M/170M [00:04&lt;00:00, 41.6MB/s]
+100%|##########| 170M/170M [00:04&lt;00:00, 42.9MB/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 &#39;trunc&#39; function NOT &#39;floor&#39;). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode=&#39;trunc&#39;), or for actual floor division, use torch.div(a, b, rounding_mode=& [...]
@@ -576,7 +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  16.971 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes  24.737 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 8e4a83a16e..089b356433 100644
--- a/docs/how_to/deploy_models/deploy_prequantized.html
+++ b/docs/how_to/deploy_models/deploy_prequantized.html
@@ -497,9 +497,7 @@ training. Other models require a full post training calibration.</p>
 Downloading: &quot;https://download.pytorch.org/models/mobilenet_v2-b0353104.pth&quot; to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2-b0353104.pth
 
   0%|          | 0.00/13.6M [00:00&lt;?, ?B/s]
- 47%|####6     | 6.30M/13.6M [00:00&lt;00:00, 63.1MB/s]
- 91%|######### | 12.3M/13.6M [00:00&lt;00:00, 44.4MB/s]
-100%|##########| 13.6M/13.6M [00:00&lt;00:00, 50.0MB/s]
+100%|##########| 13.6M/13.6M [00:00&lt;00:00, 159MB/s]
 </pre></div>
 </div>
 </div>
@@ -590,7 +588,7 @@ 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)
-  90.4050      90.3277      95.0359      90.0595       0.4947
+  90.6324      90.5561      93.1118      90.2627       0.4082
 </pre></div>
 </div>
 <div class="admonition note">
@@ -629,7 +627,7 @@ 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  6.767 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  8.512 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 2bcf0c8f62..4fd6a01496 100644
--- a/docs/how_to/deploy_models/deploy_prequantized_tflite.html
+++ b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
@@ -582,7 +582,7 @@ 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)
-  120.9329     120.8553     125.1488     120.2441      0.5669
+  121.4938     121.4546     125.6496     120.6399      0.5865
 </pre></div>
 </div>
 <div class="admonition note">
@@ -610,7 +610,7 @@ 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  31.481 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes  28.305 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 f390391acf..f15f121577 100644
--- a/docs/how_to/deploy_models/deploy_quantized.html
+++ b/docs/how_to/deploy_models/deploy_quantized.html
@@ -520,7 +520,7 @@ 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  39.807 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  42.212 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 2bc958574b..c158302807 100644
--- a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
+++ b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
@@ -462,22 +462,23 @@ 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&lt;?, ?KB/s]
-  6%|5         | 7359/132723 [00:00&lt;00:01, 73578.32KB/s]
- 12%|#2        | 16072/132723 [00:00&lt;00:01, 81546.45KB/s]
- 18%|#8        | 24227/132723 [00:00&lt;00:01, 62005.86KB/s]
- 25%|##4       | 32757/132723 [00:00&lt;00:02, 43299.33KB/s]
- 31%|###1      | 41408/132723 [00:00&lt;00:01, 53118.25KB/s]
- 38%|###7      | 50040/132723 [00:00&lt;00:01, 61252.25KB/s]
- 44%|####4     | 58670/132723 [00:00&lt;00:01, 67745.91KB/s]
- 51%|#####     | 67351/132723 [00:01&lt;00:00, 72896.78KB/s]
- 57%|#####7    | 75784/132723 [00:01&lt;00:00, 76080.18KB/s]
- 63%|######3   | 84228/132723 [00:01&lt;00:00, 78457.65KB/s]
- 70%|######9   | 92783/132723 [00:01&lt;00:00, 80506.54KB/s]
- 76%|#######6  | 101108/132723 [00:01&lt;00:00, 70659.05KB/s]
- 83%|########2 | 109815/132723 [00:01&lt;00:00, 75025.98KB/s]
- 89%|########8 | 118048/132723 [00:01&lt;00:00, 77036.58KB/s]
- 95%|#########5| 126742/132723 [00:01&lt;00:00, 79831.45KB/s]
-100%|##########| 132723/132723 [00:01&lt;00:00, 69886.02KB/s]
+  5%|4         | 6163/132723 [00:00&lt;00:02, 61623.30KB/s]
+ 11%|#1        | 14890/132723 [00:00&lt;00:01, 76706.81KB/s]
+ 17%|#6        | 22561/132723 [00:00&lt;00:01, 61488.26KB/s]
+ 24%|##3       | 31336/132723 [00:00&lt;00:01, 70502.66KB/s]
+ 29%|##9       | 39069/132723 [00:00&lt;00:01, 72748.29KB/s]
+ 36%|###6      | 47847/132723 [00:00&lt;00:01, 77552.97KB/s]
+ 42%|####2     | 55779/132723 [00:00&lt;00:01, 60308.08KB/s]
+ 49%|####8     | 64533/132723 [00:00&lt;00:01, 67303.49KB/s]
+ 54%|#####4    | 71883/132723 [00:01&lt;00:00, 68966.09KB/s]
+ 61%|######    | 80613/132723 [00:01&lt;00:00, 74045.10KB/s]
+ 67%|######7   | 89317/132723 [00:01&lt;00:00, 77730.07KB/s]
+ 74%|#######3  | 97956/132723 [00:01&lt;00:00, 80224.79KB/s]
+ 80%|########  | 106638/132723 [00:01&lt;00:00, 82146.77KB/s]
+ 87%|########6 | 115007/132723 [00:01&lt;00:00, 66534.79KB/s]
+ 93%|#########3| 123756/132723 [00:01&lt;00:00, 71833.44KB/s]
+ 99%|#########9| 131450/132723 [00:01&lt;00:00, 73184.77KB/s]
+100%|##########| 132723/132723 [00:01&lt;00:00, 71855.54KB/s]
 </pre></div>
 </div>
 <p>Create TVM runtime and do inference
@@ -516,7 +517,7 @@ Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from h
 <span class="n">plt</span><span class="o">.</span><span class="n">show</span><span class="p">()</span>
 </pre></div>
 </div>
-<img src="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" srcset="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" alt="deploy ssd gluoncv" class = "sphx-glr-single-img"/><p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes  2.439 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  7.321 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 c7dec099dd..67e6391281 100644
--- a/docs/how_to/deploy_models/sg_execution_times.html
+++ b/docs/how_to/deploy_models/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-deploy-models-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>13:04.439</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>13:21.255</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 86%" />
@@ -349,35 +349,35 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="deploy_object_detection_pytorch.html#sphx-glr-how-to-deploy-models-deploy-object-detection-pytorch-py"><span class="std std-ref">Compile PyTorch Object Detection Models</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_object_detection_pytorch.py</span></code>)</p></td>
-<td><p>03:16.971</p></td>
+<td><p>03:24.737</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="deploy_ssd_gluoncv.html#sphx-glr-how-to-deploy-models-deploy-ssd-gluoncv-py"><span class="std std-ref">Deploy Single Shot Multibox Detector(SSD) model</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_ssd_gluoncv.py</span></code>)</p></td>
-<td><p>03:02.439</p></td>
+<td><p>03:07.321</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>02:31.481</p></td>
+<td><p>02:28.305</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>01:39.807</p></td>
+<td><p>01:42.212</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>01:06.767</p></td>
+<td><p>01:08.512</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="deploy_model_on_android.html#sphx-glr-how-to-deploy-models-deploy-model-on-android-py"><span class="std std-ref">Deploy the Pretrained Model on Android</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_android.py</span></code>)</p></td>
-<td><p>00:36.121</p></td>
+<td><p>00:37.861</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="deploy_model_on_nano.html#sphx-glr-how-to-deploy-models-deploy-model-on-nano-py"><span class="std std-ref">Deploy the Pretrained Model on Jetson Nano</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_nano.py</span></code>)</p></td>
-<td><p>00:25.702</p></td>
+<td><p>00:26.511</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="deploy_model_on_rasp.html#sphx-glr-how-to-deploy-models-deploy-model-on-rasp-py"><span class="std std-ref">Deploy the Pretrained Model on Raspberry Pi</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_rasp.py</span></code>)</p></td>
-<td><p>00:25.146</p></td>
+<td><p>00:25.790</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="deploy_sparse.html#sphx-glr-how-to-deploy-models-deploy-sparse-py"><span class="std std-ref">Deploy a Hugging Face Pruned Model on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_sparse.py</span></code>)</p></td>
diff --git a/docs/how_to/extend_tvm/bring_your_own_datatypes.html b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
index de9eae7ee1..b957d82b27 100644
--- a/docs/how_to/extend_tvm/bring_your_own_datatypes.html
+++ b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
@@ -621,7 +621,7 @@ 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.zipb5ec6e0e-1a2e-4ead-b5c9-a1715a34eb53 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.zip85164043-6ab1-4ce7-9f25-a44176bc77c5 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 f076ce89ae..47425a31d8 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>00:48.729</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:49.063</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>00:45.240</p></td>
+<td><p>00:45.485</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>00:02.441</p></td>
+<td><p>00:02.505</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>00:01.040</p></td>
+<td><p>00:01.065</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 a7c187a4d4..4b42d9094f 100644
--- a/docs/how_to/extend_tvm/use_pass_instrument.html
+++ b/docs/how_to/extend_tvm/use_pass_instrument.html
@@ -525,10 +525,10 @@ 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: 7091us [7091us] (45.25%; 45.25%)
-FoldScaleAxis: 8579us [7us] (54.75%; 54.75%)
-        FoldConstant: 8572us [1775us] (54.70%; 99.92%)
-                InferType: 6796us [6796us] (43.37%; 79.29%)
+InferType: 7311us [7311us] (46.27%; 46.27%)
+FoldScaleAxis: 8490us [8us] (53.73%; 53.73%)
+        FoldConstant: 8482us [1698us] (53.68%; 99.90%)
+                InferType: 6783us [6783us] (42.93%; 79.98%)
 </pre></div>
 </div>
 </div>
@@ -550,10 +550,10 @@ 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: 6623us [6623us] (44.82%; 44.82%)
-FoldScaleAxis: 8153us [5us] (55.18%; 55.18%)
-        FoldConstant: 8148us [1692us] (55.14%; 99.94%)
-                InferType: 6456us [6456us] (43.69%; 79.24%)
+InferType: 6902us [6902us] (44.90%; 44.90%)
+FoldScaleAxis: 8470us [6us] (55.10%; 55.10%)
+        FoldConstant: 8464us [1732us] (55.06%; 99.93%)
+                InferType: 6732us [6732us] (43.79%; 79.53%)
 </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 6e6394afb1..f4c67ed43e 100644
--- a/docs/how_to/optimize_operators/opt_conv_cuda.html
+++ b/docs/how_to/optimize_operators/opt_conv_cuda.html
@@ -577,7 +577,7 @@ latency of convolution.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Convolution: </span><span class="si">%f</span><span class="s2"> ms&quot;</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: 34.213375 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 54.245216 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 772dbde193..3082283a9f 100644
--- a/docs/how_to/optimize_operators/opt_conv_tensorcore.html
+++ b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
@@ -914,7 +914,7 @@ be able to run on our build server</p>
     <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;conv2d with tensor core: </span><span class="si">%f</span><span class="s2"> ms&quot;</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: 13.353340 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 13.002118 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 c6a071a04d..588d1d4676 100644
--- a/docs/how_to/optimize_operators/opt_gemm.html
+++ b/docs/how_to/optimize_operators/opt_gemm.html
@@ -474,8 +474,8 @@ 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">&quot;Baseline: </span><span class="si">%f</span><span class="s2">&quot;</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: 0.019583
-Baseline: 3.429270
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.019491
+Baseline: 3.290225
 </pre></div>
 </div>
 <p>In TVM, we can always inspect lower level IR to debug or optimize our schedule.
@@ -534,7 +534,7 @@ 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">&quot;Opt1: </span><span class="si">%f</span><span class="s2">&quot;</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: 0.308873
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.328238
 </pre></div>
 </div>
 <p>Here is the generated IR after blocking.</p>
@@ -600,7 +600,7 @@ vastly.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt2: </span><span class="si">%f</span><span class="s2">&quot;</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: 0.331130
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.353639
 </pre></div>
 </div>
 <p>Here is the generated IR after vectorization.</p>
@@ -660,7 +660,7 @@ the access pattern for A matrix is more cache friendly.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt3: </span><span class="si">%f</span><span class="s2">&quot;</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: 0.116499
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.122369
 </pre></div>
 </div>
 <p>Here is the generated IR after loop permutation.</p>
@@ -742,7 +742,7 @@ flattening.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt4: </span><span class="si">%f</span><span class="s2">&quot;</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: 0.109366
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109637
 </pre></div>
 </div>
 <p>Here is the generated IR after array packing.</p>
@@ -827,7 +827,7 @@ write to C when all the block results are ready.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt5: </span><span class="si">%f</span><span class="s2">&quot;</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: 0.111082
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.111313
 </pre></div>
 </div>
 <p>Here is the generated IR after blocking.</p>
@@ -916,7 +916,7 @@ write to C when all the block results are ready.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt6: </span><span class="si">%f</span><span class="s2">&quot;</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: 0.147840
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.147555
 </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 bf2257e4d6..d32706e36a 100644
--- a/docs/how_to/optimize_operators/sg_execution_times.html
+++ b/docs/how_to/optimize_operators/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-optimize-operators-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:35.433</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:35.420</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 83%" />
@@ -349,15 +349,15 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="opt_gemm.html#sphx-glr-how-to-optimize-operators-opt-gemm-py"><span class="std std-ref">How to optimize GEMM on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_gemm.py</span></code>)</p></td>
-<td><p>00:32.596</p></td>
+<td><p>00:32.731</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>00:01.698</p></td>
+<td><p>00:01.512</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>00:01.138</p></td>
+<td><p>00:01.176</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 d1caedd9db..99d6075a32 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>09:22.962</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>09:16.506</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>05:52.239</p></td>
+<td><p>05:43.754</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>01:34.879</p></td>
+<td><p>01:34.813</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>01:04.373</p></td>
+<td><p>01:04.752</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>00:27.840</p></td>
+<td><p>00:28.883</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>00:12.218</p></td>
+<td><p>00:12.495</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>00:11.414</p></td>
+<td><p>00:11.809</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 4657ab6747..8df73536b8 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
@@ -503,191 +503,952 @@ cooperative fetching, unrolling and operator fusion.</p>
              bias: Buffer(bias_2: Pointer(float32), float32, [1, 512, 1, 1], []),
              compute: Buffer(compute_2: Pointer(float32), float32, [1, 512, 7, 7], [])}
   buffer_map = {data_1: data, kernel_1: kernel, bias_1: bias, compute_1: compute} {
-  attr [IterVar(blockIdx.x: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;blockIdx.x&quot;)] &quot;thread_extent&quot; = 128;
-  allocate(conv2d_nchw: Pointer(local float32), float32, [4]), storage_scope = local;
-  allocate(pad_temp.shared: Pointer(shared float32), float32, [324]), storage_scope = shared;
-  allocate(kernel.shared: Pointer(shared float32), float32, [144]), storage_scope = shared;
+  attr [IterVar(blockIdx.x: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;blockIdx.x&quot;)] &quot;thread_extent&quot; = 32;
+  allocate(conv2d_nchw: Pointer(local float32), float32, [16]), storage_scope = local;
+  allocate(pad_temp.shared: Pointer(shared float32), float32, [784]), storage_scope = shared;
+  allocate(kernel.shared: Pointer(shared float32), float32, [256]), storage_scope = shared;
   attr [IterVar(threadIdx.x: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49 {
-    conv2d_nchw_1: Buffer(conv2d_nchw, float32, [4], [], scope=&quot;local&quot;, align=16)[0] = 0f32
-    conv2d_nchw_1[1] = 0f32
+    conv2d_nchw_1: Buffer(conv2d_nchw, float32, [4], [], scope=&quot;local&quot;, align=8)[0] = 0f32
     conv2d_nchw_1[2] = 0f32
+    conv2d_nchw_1[4] = 0f32
+    conv2d_nchw_1[6] = 0f32
+    conv2d_nchw_1[8] = 0f32
+    conv2d_nchw_1[10] = 0f32
+    conv2d_nchw_1[12] = 0f32
+    conv2d_nchw_1[14] = 0f32
+    conv2d_nchw_1[1] = 0f32
     conv2d_nchw_1[3] = 0f32
-    for (rc.outer.outer: int32, 0, 128) {
-      let cse_var_2: int32 = (rc.outer.outer*196)
-      let cse_var_1: int32 = (rc.outer.outer*36)
-       {
+    conv2d_nchw_1[5] = 0f32
+    conv2d_nchw_1[7] = 0f32
+    conv2d_nchw_1[9] = 0f32
+    conv2d_nchw_1[11] = 0f32
+    conv2d_nchw_1[13] = 0f32
+    conv2d_nchw_1[15] = 0f32
+    for (rc.outer.outer: int32, 0, 32) {
+      for (ry.outer.outer: int32, 0, 3) {
         attr [IterVar(threadIdx.x_1: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        pad_temp.shared_1: Buffer(pad_temp.shared, float32, [324], [], scope=&quot;shared&quot;)[threadIdx.x_1] = @tir.if_then_else((((9 &lt;= threadIdx.x_1) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 9))) &amp;&amp; (floormod(threadIdx.x_1, 9) &lt; 8)), data_3: Buffer(data_2, float32, [25088], [])[(((cse_var_2 + (floordiv(threadIdx.x_1, 9)*7)) + floormod(threadIdx.x_1, 9)) - 8)], 0f32, dtype=float32)
+        pad_temp.shared_1: Buffer(pad_temp.shared, float32, [784], [], scope=&quot;shared&quot;)[threadIdx.x_1] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3: Buffer(data_2, float32, [25088], [])[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) - 8)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 41)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 90)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 139)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 188)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 237)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 286)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 343)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 335)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 392)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 384)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 441)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 433)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 490)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 482)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 539)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 531)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 588)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 580)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 637)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 629)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 686)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 678)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 735)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 7))), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 727)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_2: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49 {
+          kernel.shared_1: Buffer(kernel.shared, float32, [256], [], scope=&quot;shared&quot;)[(threadIdx.x_2*4)] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[(((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3))]
+          kernel.shared_1[((threadIdx.x_2*4) + 1)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 9)]
+          kernel.shared_1[((threadIdx.x_2*4) + 2)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 18)]
+          kernel.shared_1[((threadIdx.x_2*4) + 3)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 27)]
+        }
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        if @tir.likely((threadIdx.x_2 &lt; 15), dtype=bool) {
+          let cse_var_1: int32 = (ry.outer.outer*3)
+           {
+            kernel.shared_1[((threadIdx.x_2*4) + 196)] = kernel_3[(((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 4), 16)*9)) + cse_var_1)]
+            kernel.shared_1[((threadIdx.x_2*4) + 197)] = kernel_3[(((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 5), 16)*9)) + cse_var_1)]
+            kernel.shared_1[((threadIdx.x_2*4) + 198)] = kernel_3[(((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 6), 16)*9)) + cse_var_1)]
+            kernel.shared_1[((threadIdx.x_2*4) + 199)] = kernel_3[(((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 7), 16)*9)) + cse_var_1)]
+          }
+        }
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[0]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[32]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[64]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[96]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[128]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[160]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[192]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[224]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[16]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[48]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[80]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[112]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[144]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[176]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[208]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[240]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[1]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[33]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[65]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[97]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[129]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[161]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[193]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[225]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[17]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[49]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[81]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[113]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[145]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[177]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[209]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[241]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[2]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[34]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[66]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[98]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[130]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[162]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[194]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[226]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[18]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[50]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[82]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[114]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[146]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[178]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[210]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[242]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[3]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[35]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[67]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[99]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[131]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[163]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[195]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[227]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[19]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[51]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[83]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[115]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[147]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[179]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[211]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[243]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[4]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[36]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[68]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[100]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[132]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[164]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[196]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[228]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[20]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[52]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[84]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[116]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[148]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[180]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[212]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[244]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[5]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[37]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[69]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[101]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[133]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[165]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[197]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[229]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[21]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[53]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[85]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[117]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[149]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[181]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[213]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[245]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[6]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[38]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[70]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[102]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[134]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[166]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[198]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[230]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[22]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[54]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[86]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[118]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[150]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[182]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[214]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[246]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[7]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[39]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[71]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[103]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[135]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[167]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[199]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[231]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[23]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[55]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[87]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[119]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[151]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[183]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[215]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[247]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[8]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[40]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[72]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[104]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[136]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[168]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[200]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[232]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[24]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[56]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[88]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[120]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[152]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[184]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[216]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[248]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[9]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[41]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[73]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[105]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[137]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[169]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[201]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[233]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[25]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[57]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[89]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[121]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[153]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[185]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[217]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[249]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[10]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[42]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[74]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[106]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[138]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[170]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[202]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[234]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[26]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[58]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[90]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[122]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[154]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[186]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[218]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[250]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[11]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[43]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[75]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[107]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[139]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[171]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[203]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[235]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[27]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[59]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[91]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[123]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[155]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[187]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[219]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[251]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[12]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[44]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[76]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[108]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[140]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[172]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[204]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[236]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[28]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[60]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[92]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[124]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[156]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[188]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[220]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[252]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[13]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[45]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[77]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[109]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[141]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[173]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[205]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[237]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[29]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[61]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[93]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[125]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[157]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[189]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[221]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[253]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[14]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[46]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[78]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[110]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[142]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[174]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[206]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[238]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[30]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[62]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[94]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[126]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[158]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[190]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[222]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[254]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[15]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[47]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[79]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[111]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[143]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[175]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[207]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[239]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[31]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[63]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[95]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[127]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[159]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[191]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[223]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[255]))
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[threadIdx.x_1] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) - 7)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 42)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 91)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 140)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 189)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 238)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 287)], 0f32, dtype=float32)
         attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else(((((9 &lt;= floormod((threadIdx.x_1 + 49), 81)) &amp;&amp; (floormod((threadIdx.x_1 + 49), 81) &lt; 72)) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1 + 4), 9))) &amp;&amp; (floormod((threadIdx.x_1 + 4), 9) &lt; 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 49), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 49), 81), 9)*7)) + floormod((threadIdx.x_1 + 4), 9)) - 8)], 0f32, dtype=float32)
+        pad_temp.shared_1[(threadIdx.x_1 + 343)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 336)], 0f32, dtype=float32)
         attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else(((1 &lt;= floormod((threadIdx.x_1 + 8), 9)) &amp;&amp; (floormod((threadIdx.x_1 + 8), 9) &lt; 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 98), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 17), 81), 9)*7)) + floormod((threadIdx.x_1 + 8), 9)) - 8)], 0f32, dtype=float32)
+        pad_temp.shared_1[(threadIdx.x_1 + 392)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 385)], 0f32, dtype=float32)
         attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else(((((9 &lt;= floormod((threadIdx.x_1 + 66), 81)) &amp;&amp; (floormod((threadIdx.x_1 + 66), 81) &lt; 72)) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1 + 3), 9))) &amp;&amp; (floormod((threadIdx.x_1 + 3), 9) &lt; 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 147), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 66), 81), 9)*7)) + floormod((threadIdx.x_1 + 3), 9)) - 8)], 0f32, dtype=float32)
+        pad_temp.shared_1[(threadIdx.x_1 + 441)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 434)], 0f32, dtype=float32)
         attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else(((((9 &lt;= floormod((threadIdx.x_1 + 34), 81)) &amp;&amp; (floormod((threadIdx.x_1 + 34), 81) &lt; 72)) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1 + 7), 9))) &amp;&amp; (floormod((threadIdx.x_1 + 7), 9) &lt; 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 196), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 34), 81), 9)*7)) + floormod((threadIdx.x_1 + 7), 9)) - 8)], 0f32, dtype=float32)
+        pad_temp.shared_1[(threadIdx.x_1 + 490)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 483)], 0f32, dtype=float32)
         attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else((((9 &lt;= floormod((threadIdx.x_1 + 2), 81)) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1 + 2), 9))) &amp;&amp; (floormod((threadIdx.x_1 + 2), 9) &lt; 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 245), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 2), 81), 9)*7)) + floormod((threadIdx.x_1 + 2), 9)) - 8)], 0f32, dtype=float32)
+        pad_temp.shared_1[(threadIdx.x_1 + 539)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 532)], 0f32, dtype=float32)
         attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        if @tir.likely((threadIdx.x_1 &lt; 30), dtype=bool) {
-          pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else((((threadIdx.x_1 &lt; 21) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1 + 6), 9))) &amp;&amp; (floormod((threadIdx.x_1 + 6), 9) &lt; 8)), data_3[((((cse_var_2 + (floordiv((threadIdx.x_1 + 294), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 51), 81), 9)*7)) + floormod((threadIdx.x_1 + 6), 9)) - 8)], 0f32, dtype=float32)
+        pad_temp.shared_1[(threadIdx.x_1 + 588)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 581)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 637)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 630)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 686)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 679)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 735)] = @tir.if_then_else(((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 728)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49 {
+          kernel.shared_1[(threadIdx.x_2*4)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 1)]
+          kernel.shared_1[((threadIdx.x_2*4) + 1)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 10)]
+          kernel.shared_1[((threadIdx.x_2*4) + 2)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 19)]
+          kernel.shared_1[((threadIdx.x_2*4) + 3)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 28)]
         }
-        attr [IterVar(threadIdx.x_2: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        kernel.shared_1: Buffer(kernel.shared, float32, [144], [], scope=&quot;shared&quot;)[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[((((blockIdx.x*18432) + (floordiv(threadIdx.x_2, 36)*4608)) + cse_var_1) + floormod(threadIdx.x_2, 36))]
         attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        kernel.shared_1[(threadIdx.x_2 + 49)] = kernel_3[((((((blockIdx.x*18432) + (floordiv((threadIdx.x_2 + 49), 36)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 13), 36), 9)*9)) + (floordiv(floormod((threadIdx.x_2 + 4), 9), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        if @tir.likely((threadIdx.x_2 &lt; 15), dtype=bool) {
+          let cse_var_2: int32 = (ry.outer.outer*3)
+           {
+            kernel.shared_1[((threadIdx.x_2*4) + 196)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 4), 16)*9)) + cse_var_2) + 1)]
+            kernel.shared_1[((threadIdx.x_2*4) + 197)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 5), 16)*9)) + cse_var_2) + 1)]
+            kernel.shared_1[((threadIdx.x_2*4) + 198)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 6), 16)*9)) + cse_var_2) + 1)]
+            kernel.shared_1[((threadIdx.x_2*4) + 199)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 7), 16)*9)) + cse_var_2) + 1)]
+          }
+        }
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[0]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[32]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[64]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[96]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[128]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[160]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[192]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[224]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[16]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[48]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[80]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[112]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[144]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[176]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[208]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[240]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[1]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[33]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[65]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[97]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[129]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[161]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[193]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[225]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[17]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[49]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[81]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[113]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[145]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[177]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[209]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[241]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[2]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[34]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[66]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[98]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[130]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[162]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[194]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[226]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[18]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[50]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[82]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[114]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[146]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[178]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[210]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[242]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[3]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[35]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[67]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[99]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[131]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[163]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[195]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[227]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[19]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[51]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[83]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[115]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[147]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[179]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[211]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[243]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[4]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[36]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[68]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[100]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[132]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[164]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[196]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[228]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[20]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[52]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[84]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[116]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[148]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[180]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[212]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[244]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[5]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[37]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[69]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[101]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[133]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[165]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[197]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[229]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[21]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[53]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[85]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[117]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[149]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[181]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[213]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[245]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[6]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[38]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[70]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[102]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[134]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[166]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[198]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[230]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[22]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[54]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[86]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[118]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[150]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[182]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[214]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[246]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[7]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[39]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[71]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[103]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[135]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[167]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[199]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[231]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[23]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[55]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[87]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[119]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[151]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[183]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[215]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[247]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[8]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[40]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[72]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[104]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[136]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[168]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[200]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[232]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[24]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[56]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[88]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[120]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[152]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[184]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[216]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[248]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[9]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[41]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[73]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[105]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[137]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[169]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[201]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[233]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[25]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[57]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[89]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[121]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[153]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[185]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[217]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[249]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[10]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[42]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[74]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[106]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[138]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[170]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[202]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[234]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[26]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[58]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[90]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[122]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[154]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[186]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[218]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[250]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[11]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[43]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[75]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[107]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[139]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[171]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[203]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[235]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[27]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[59]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[91]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[123]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[155]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[187]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[219]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[251]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[12]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[44]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[76]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[108]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[140]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[172]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[204]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[236]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[28]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[60]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[92]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[124]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[156]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[188]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[220]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[252]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[13]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[45]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[77]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[109]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[141]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[173]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[205]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[237]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[29]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[61]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[93]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[125]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[157]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[189]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[221]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[253]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[14]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[46]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[78]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[110]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[142]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[174]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[206]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[238]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[30]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[62]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[94]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[126]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[158]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[190]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[222]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[254]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[15]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[47]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[79]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[111]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[143]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[175]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[207]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[239]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[31]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[63]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[95]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[127]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[159]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[191]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[223]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[255]))
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[threadIdx.x_1] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) - 6)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 43)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 92)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 141)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 190)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 239)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 288)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 343)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 337)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 392)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 386)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 441)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 435)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 490)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 484)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 539)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 533)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 588)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 582)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 637)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 631)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 686)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 680)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
+        pad_temp.shared_1[(threadIdx.x_1 + 735)] = @tir.if_then_else((((1 &lt;= (floordiv(threadIdx.x_1, 7) + ry.outer.outer)) &amp;&amp; ((floordiv(threadIdx.x_1, 7) + ry.outer.outer) &lt; 8)) &amp;&amp; (floormod(threadIdx.x_1, 7) &lt; 6)), data_3[((((rc.outer.outer*784) + (ry.outer.outer*7)) + threadIdx.x_1) + 729)], 0f32, dtype=float32)
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49 {
+          kernel.shared_1[(threadIdx.x_2*4)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 2)]
+          kernel.shared_1[((threadIdx.x_2*4) + 1)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 11)]
+          kernel.shared_1[((threadIdx.x_2*4) + 2)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 20)]
+          kernel.shared_1[((threadIdx.x_2*4) + 3)] = kernel_3[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 4)*4608)) + (rc.outer.outer*144)) + (floormod(threadIdx.x_2, 4)*36)) + (ry.outer.outer*3)) + 29)]
+        }
         attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 49;
-        if @tir.likely((threadIdx.x_2 &lt; 46), dtype=bool) {
-          kernel.shared_1[(threadIdx.x_2 + 98)] = kernel_3[((((((blockIdx.x*18432) + (floordiv((threadIdx.x_2 + 98), 36)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 26), 36), 9)*9)) + (floordiv(floormod((threadIdx.x_2 + 8), 9), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        if @tir.likely((threadIdx.x_2 &lt; 15), dtype=bool) {
+          let cse_var_3: int32 = (ry.outer.outer*3)
+           {
+            kernel.shared_1[((threadIdx.x_2*4) + 196)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 4), 16)*9)) + cse_var_3) + 2)]
+            kernel.shared_1[((threadIdx.x_2*4) + 197)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 5), 16)*9)) + cse_var_3) + 2)]
+            kernel.shared_1[((threadIdx.x_2*4) + 198)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 6), 16)*9)) + cse_var_3) + 2)]
+            kernel.shared_1[((threadIdx.x_2*4) + 199)] = kernel_3[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 4)*4608)) + (rc.outer.outer*144)) + (floormod(((threadIdx.x_2*4) + 7), 16)*9)) + cse_var_3) + 2)]
+          }
         }
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7))]*kernel.shared_1[0]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[9]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7))]*kernel.shared_1[36]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[45]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7))]*kernel.shared_1[72]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[81]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7))]*kernel.shared_1[108]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[117]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[1]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 82)]*kernel.shared_1[10]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[37]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 82)]*kernel.shared_1[46]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[73]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 82)]*kernel.shared_1[82]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[109]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 82)]*kernel.shared_1[118]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[2]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 83)]*kernel.shared_1[11]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[38]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 83)]*kernel.shared_1[47]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[74]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 83)]*kernel.shared_1[83]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[110]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 83)]*kernel.shared_1[119]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[3]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[12]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[39]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[48]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[75]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[84]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[111]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[120]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 10)]*kernel.shared_1[4]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 91)]*kernel.shared_1[13]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 10)]*kernel.shared_1[40]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 91)]*kernel.shared_1[49]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 10)]*kernel.shared_1[76]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 91)]*kernel.shared_1[85]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 10)]*kernel.shared_1[112]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 91)]*kernel.shared_1[121]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 11)]*kernel.shared_1[5]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 92)]*kernel.shared_1[14]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 11)]*kernel.shared_1[41]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 92)]*kernel.shared_1[50]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 11)]*kernel.shared_1[77]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 92)]*kernel.shared_1[86]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 11)]*kernel.shared_1[113]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 92)]*kernel.shared_1[122]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[6]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[15]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[42]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[51]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[78]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[87]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[114]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[123]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 19)]*kernel.shared_1[7]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 100)]*kernel.shared_1[16]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 19)]*kernel.shared_1[43]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 100)]*kernel.shared_1[52]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 19)]*kernel.shared_1[79]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 100)]*kernel.shared_1[88]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 19)]*kernel.shared_1[115]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 100)]*kernel.shared_1[124]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 20)]*kernel.shared_1[8]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 101)]*kernel.shared_1[17]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 20)]*kernel.shared_1[44]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 101)]*kernel.shared_1[53]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 20)]*kernel.shared_1[80]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 101)]*kernel.shared_1[89]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 20)]*kernel.shared_1[116]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 101)]*kernel.shared_1[125]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[18]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[27]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[54]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[63]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[90]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[99]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[126]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[135]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 163)]*kernel.shared_1[19]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 244)]*kernel.shared_1[28]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 163)]*kernel.shared_1[55]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 244)]*kernel.shared_1[64]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 163)]*kernel.shared_1[91]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 244)]*kernel.shared_1[100]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 163)]*kernel.shared_1[127]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 244)]*kernel.shared_1[136]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 164)]*kernel.shared_1[20]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 245)]*kernel.shared_1[29]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 164)]*kernel.shared_1[56]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 245)]*kernel.shared_1[65]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 164)]*kernel.shared_1[92]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 245)]*kernel.shared_1[101]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 164)]*kernel.shared_1[128]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 245)]*kernel.shared_1[137]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[21]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[30]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[57]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[66]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[93]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[102]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[129]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[138]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 172)]*kernel.shared_1[22]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 253)]*kernel.shared_1[31]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 172)]*kernel.shared_1[58]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 253)]*kernel.shared_1[67]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 172)]*kernel.shared_1[94]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 253)]*kernel.shared_1[103]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 172)]*kernel.shared_1[130]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 253)]*kernel.shared_1[139]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 173)]*kernel.shared_1[23]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 254)]*kernel.shared_1[32]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 173)]*kernel.shared_1[59]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 254)]*kernel.shared_1[68]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 173)]*kernel.shared_1[95]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 254)]*kernel.shared_1[104]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 173)]*kernel.shared_1[131]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 254)]*kernel.shared_1[140]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[24]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[33]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[60]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[69]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[96]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[105]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[132]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[141]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 181)]*kernel.shared_1[25]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 262)]*kernel.shared_1[34]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 181)]*kernel.shared_1[61]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 262)]*kernel.shared_1[70]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 181)]*kernel.shared_1[97]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 262)]*kernel.shared_1[106]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 181)]*kernel.shared_1[133]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 262)]*kernel.shared_1[142]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 182)]*kernel.shared_1[26]))
-        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 263)]*kernel.shared_1[35]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 182)]*kernel.shared_1[62]))
-        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 263)]*kernel.shared_1[71]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 182)]*kernel.shared_1[98]))
-        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 263)]*kernel.shared_1[107]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 182)]*kernel.shared_1[134]))
-        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + floormod(threadIdx.x, 7)) + 263)]*kernel.shared_1[143]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[0]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[32]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[64]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[96]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[128]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[160]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[192]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[224]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[16]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[48]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[80]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[112]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[144]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[176]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[208]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[threadIdx.x]*kernel.shared_1[240]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[1]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[33]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[65]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[97]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[129]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[161]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[193]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[225]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[17]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[49]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[81]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[113]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[145]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[177]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[209]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 49)]*kernel.shared_1[241]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[2]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[34]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[66]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[98]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[130]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[162]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[194]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[226]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[18]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[50]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[82]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[114]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[146]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[178]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[210]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 98)]*kernel.shared_1[242]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[3]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[35]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[67]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[99]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[131]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[163]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[195]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[227]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[19]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[51]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[83]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[115]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[147]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[179]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[211]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 147)]*kernel.shared_1[243]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[4]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[36]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[68]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[100]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[132]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[164]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[196]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[228]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[20]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[52]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[84]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[116]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[148]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[180]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[212]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 196)]*kernel.shared_1[244]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[5]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[37]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[69]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[101]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[133]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[165]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[197]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[229]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[21]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[53]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[85]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[117]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[149]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[181]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[213]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 245)]*kernel.shared_1[245]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[6]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[38]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[70]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[102]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[134]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[166]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[198]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[230]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[22]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[54]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[86]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[118]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[150]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[182]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[214]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 294)]*kernel.shared_1[246]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[7]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[39]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[71]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[103]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[135]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[167]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[199]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[231]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[23]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[55]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[87]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[119]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[151]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[183]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[215]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 343)]*kernel.shared_1[247]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[8]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[40]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[72]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[104]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[136]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[168]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[200]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[232]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[24]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[56]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[88]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[120]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[152]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[184]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[216]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 392)]*kernel.shared_1[248]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[9]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[41]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[73]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[105]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[137]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[169]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[201]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[233]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[25]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[57]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[89]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[121]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[153]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[185]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[217]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 441)]*kernel.shared_1[249]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[10]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[42]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[74]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[106]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[138]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[170]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[202]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[234]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[26]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[58]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[90]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[122]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[154]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[186]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[218]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 490)]*kernel.shared_1[250]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[11]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[43]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[75]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[107]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[139]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[171]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[203]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[235]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[27]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[59]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[91]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[123]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[155]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[187]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[219]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 539)]*kernel.shared_1[251]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[12]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[44]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[76]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[108]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[140]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[172]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[204]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[236]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[28]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[60]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[92]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[124]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[156]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[188]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[220]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 588)]*kernel.shared_1[252]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[13]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[45]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[77]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[109]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[141]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[173]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[205]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[237]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[29]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[61]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[93]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[125]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[157]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[189]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[221]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 637)]*kernel.shared_1[253]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[14]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[46]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[78]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[110]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[142]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[174]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[206]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[238]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[30]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[62]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[94]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[126]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[158]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[190]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[222]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 686)]*kernel.shared_1[254]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[15]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[47]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[79]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[111]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[143]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[175]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[207]))
+        conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[239]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[31]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[63]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[95]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[127]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[159]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[191]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[223]))
+        conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(threadIdx.x + 735)]*kernel.shared_1[255]))
       }
     }
-    for (i1.inner: int32, 0, 4) {
-      compute_3: Buffer(compute_2, float32, [25088], [])[(((blockIdx.x*196) + (i1.inner*49)) + threadIdx.x)] = max((conv2d_nchw_1[i1.inner] + bias_3: Buffer(bias_2, float32, [512], [])[((blockIdx.x*4) + i1.inner)]), 0f32)
+    for (i1.inner: int32, 0, 2) {
+      compute_3: Buffer(compute_2, float32, [25088], [])[(((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x)] = max((conv2d_nchw_1[i1.inner] + bias_3: Buffer(bias_2, float32, [512], [])[((blockIdx.x*16) + i1.inner)]), 0f32)
+      compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 98)] = max((conv2d_nchw_1[(i1.inner + 2)] + bias_3[(((blockIdx.x*16) + i1.inner) + 2)]), 0f32)
+      compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 196)] = max((conv2d_nchw_1[(i1.inner + 4)] + bias_3[(((blockIdx.x*16) + i1.inner) + 4)]), 0f32)
+      compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 294)] = max((conv2d_nchw_1[(i1.inner + 6)] + bias_3[(((blockIdx.x*16) + i1.inner) + 6)]), 0f32)
+      compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 392)] = max((conv2d_nchw_1[(i1.inner + 8)] + bias_3[(((blockIdx.x*16) + i1.inner) + 8)]), 0f32)
+      compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 490)] = max((conv2d_nchw_1[(i1.inner + 10)] + bias_3[(((blockIdx.x*16) + i1.inner) + 10)]), 0f32)
+      compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 588)] = max((conv2d_nchw_1[(i1.inner + 12)] + bias_3[(((blockIdx.x*16) + i1.inner) + 12)]), 0f32)
+      compute_3[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 686)] = max((conv2d_nchw_1[(i1.inner + 14)] + bias_3[(((blockIdx.x*16) + i1.inner) + 14)]), 0f32)
     }
   }
 }
@@ -724,7 +1485,7 @@ cooperative fetching, unrolling and operator fusion.</p>
 <span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.259 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.355 ms
 </pre></div>
 </div>
 </div>
@@ -753,10 +1514,10 @@ conv2d_nchw_nn_o_i, conv2d_nchw_nn_i = s[conv2d_nchw].split(conv2d_nchw_nn, fact
 conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_i, factor=1)
 conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
 conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
-conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=1)
-conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=4)
+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=1)
-conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
+conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=8)
 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)
@@ -765,19 +1526,19 @@ conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, fact
 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=2)
+conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=8)
 conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=2)
 conv2d_nchw_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_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
 conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=1)
-conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=3)
+conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=1)
 s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2d_nc [...]
 compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
 compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
 compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
-compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=4)
+compute_i1_o_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=1)
-compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=8)
 compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
 compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
 compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
@@ -800,7 +1561,7 @@ s[compute].bind(compute_i0_o_o_i_i1_o_o_i_fused_i2_o_o_i_fused_i3_o_o_i_fused, t
 compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused = s[compute].fuse(compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i)
 s[compute].bind(compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused, te.thread_axis(&quot;threadIdx.x&quot;))
 kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[kernel_shared].fuse(kernel_shared_ax0, kernel_shared_ax1, kernel_shared_ax2, kernel_shared_ax3)
-kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
+kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=4)
 s[kernel_shared].vectorize(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
 kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=49)
 s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis(&quot;threadIdx.x&quot;))
@@ -809,7 +1570,7 @@ pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fus
 s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
 pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=49)
 s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis(&quot;threadIdx.x&quot;))
-s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, &quot;auto_unroll_max_step&quot;, 512)
+s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, &quot;auto_unroll_max_step&quot;, 1024)
 s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, &quot;unroll_explicit&quot;, True)
 
 CUDA source code:
@@ -828,177 +1589,890 @@ CUDA source code:
   #define uint64_t unsigned long long
 #endif
 extern &quot;C&quot; __global__ void __launch_bounds__(49) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
-  float conv2d_nchw[4];
-  __shared__ float pad_temp_shared[324];
-  __shared__ float kernel_shared[144];
+  float conv2d_nchw[16];
+  __shared__ float pad_temp_shared[784];
+  __shared__ float kernel_shared[256];
   conv2d_nchw[0] = 0.000000e+00f;
-  conv2d_nchw[1] = 0.000000e+00f;
   conv2d_nchw[2] = 0.000000e+00f;
+  conv2d_nchw[4] = 0.000000e+00f;
+  conv2d_nchw[6] = 0.000000e+00f;
+  conv2d_nchw[8] = 0.000000e+00f;
+  conv2d_nchw[10] = 0.000000e+00f;
+  conv2d_nchw[12] = 0.000000e+00f;
+  conv2d_nchw[14] = 0.000000e+00f;
+  conv2d_nchw[1] = 0.000000e+00f;
   conv2d_nchw[3] = 0.000000e+00f;
-  for (int rc_outer_outer = 0; rc_outer_outer &lt; 128; ++rc_outer_outer) {
-    __syncthreads();
-    pad_temp_shared[((int)threadIdx.x)] = ((((9 &lt;= ((int)threadIdx.x)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 9))) &amp;&amp; ((((int)threadIdx.x) % 9) &lt; 8)) ? data[((((rc_outer_outer * 196) + ((((int)threadIdx.x) / 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
-    pad_temp_shared[(((int)threadIdx.x) + 49)] = (((((9 &lt;= ((((int)threadIdx.x) + 49) % 81)) &amp;&amp; (((((int)threadIdx.x) + 49) % 81) &lt; 72)) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) + 4) % 9))) &amp;&amp; (((((int)threadIdx.x) + 4) % 9) &lt; 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 49) / 81) * 49)) + ((((((int)threadIdx.x) + 49) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
-    pad_temp_shared[(((int)threadIdx.x) + 98)] = (((1 &lt;= ((((int)threadIdx.x) + 8) % 9)) &amp;&amp; (((((int)threadIdx.x) + 8) % 9) &lt; 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 98) / 81) * 49)) + (((((int)threadIdx.x) + 17) / 9) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
-    pad_temp_shared[(((int)threadIdx.x) + 147)] = (((((9 &lt;= ((((int)threadIdx.x) + 66) % 81)) &amp;&amp; (((((int)threadIdx.x) + 66) % 81) &lt; 72)) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) + 3) % 9))) &amp;&amp; (((((int)threadIdx.x) + 3) % 9) &lt; 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 147) / 81) * 49)) + ((((((int)threadIdx.x) + 66) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
-    pad_temp_shared[(((int)threadIdx.x) + 196)] = (((((9 &lt;= ((((int)threadIdx.x) + 34) % 81)) &amp;&amp; (((((int)threadIdx.x) + 34) % 81) &lt; 72)) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) + 7) % 9))) &amp;&amp; (((((int)threadIdx.x) + 7) % 9) &lt; 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 196) / 81) * 49)) + ((((((int)threadIdx.x) + 34) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] : 0.000000e+00f);
-    pad_temp_shared[(((int)threadIdx.x) + 245)] = ((((7 &lt;= ((int)threadIdx.x)) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) + 2) % 9))) &amp;&amp; (((((int)threadIdx.x) + 2) % 9) &lt; 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 245) / 81) * 49)) + (((((int)threadIdx.x) + 2) / 9) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
-    if (((int)threadIdx.x) &lt; 30) {
-      pad_temp_shared[(((int)threadIdx.x) + 294)] = ((((((int)threadIdx.x) &lt; 21) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) + 6) % 9))) &amp;&amp; (((((int)threadIdx.x) + 6) % 9) &lt; 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 294) / 81) * 49)) + (((((int)threadIdx.x) + 51) / 9) * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0.000000e+00f);
-    }
-    kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) * 18432) + ((((int)threadIdx.x) / 36) * 4608)) + (rc_outer_outer * 36)) + (((int)threadIdx.x) % 36))];
-    kernel_shared[(((int)threadIdx.x) + 49)] = kernel[((((((((int)blockIdx.x) * 18432) + (((((int)threadIdx.x) + 49) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((((int)threadIdx.x) + 13) % 36) / 9) * 9)) + ((((((int)threadIdx.x) + 4) % 9) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-    if (((int)threadIdx.x) &lt; 46) {
-      kernel_shared[(((int)threadIdx.x) + 98)] = kernel[((((((((int)blockIdx.x) * 18432) + (((((int)threadIdx.x) + 98) / 36) * 4608)) + (rc_outer_outer * 36)) + ((((((int)threadIdx.x) + 26) % 36) / 9) * 9)) + ((((((int)threadIdx.x) + 8) % 9) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+  conv2d_nchw[5] = 0.000000e+00f;
+  conv2d_nchw[7] = 0.000000e+00f;
+  conv2d_nchw[9] = 0.000000e+00f;
+  conv2d_nchw[11] = 0.000000e+00f;
+  conv2d_nchw[13] = 0.000000e+00f;
+  conv2d_nchw[15] = 0.000000e+00f;
+  for (int rc_outer_outer = 0; rc_outer_outer &lt; 32; ++rc_outer_outer) {
+    for (int ry_outer_outer = 0; ry_outer_outer &lt; 3; ++ry_outer_outer) {
+      __syncthreads();
+      pad_temp_shared[((int)threadIdx.x)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) - 8)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 49)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 41)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 98)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 90)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 147)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 139)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 196)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 188)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 245)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 237)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 294)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 286)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 343)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 335)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 392)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 384)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 441)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 433)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 490)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 482)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 539)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 531)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 588)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 580)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 637)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 629)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 686)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 678)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 735)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 7))) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 727)] : 0.000000e+00f);
+      kernel_shared[(((int)threadIdx.x) * 4)] = kernel[(((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3))];
+      kernel_shared[((((int)threadIdx.x) * 4) + 1)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 9)];
+      kernel_shared[((((int)threadIdx.x) * 4) + 2)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 18)];
+      kernel_shared[((((int)threadIdx.x) * 4) + 3)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 27)];
+      if (((int)threadIdx.x) &lt; 15) {
+        kernel_shared[((((int)threadIdx.x) * 4) + 196)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 4) &amp; 15) * 9)) + (ry_outer_outer * 3))];
+        kernel_shared[((((int)threadIdx.x) * 4) + 197)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 5) &amp; 15) * 9)) + (ry_outer_outer * 3))];
+        kernel_shared[((((int)threadIdx.x) * 4) + 198)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 6) &amp; 15) * 9)) + (ry_outer_outer * 3))];
+        kernel_shared[((((int)threadIdx.x) * 4) + 199)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 7) &amp; 15) * 9)) + (ry_outer_outer * 3))];
+      }
+      __syncthreads();
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[0]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[32]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[64]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[96]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[128]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[160]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[192]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[224]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[16]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[48]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[80]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[112]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[144]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[176]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[208]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[240]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[1]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[33]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[65]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[97]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[129]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[161]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[193]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[225]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[17]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[49]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[81]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[113]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[145]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[177]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[209]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[241]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[2]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[34]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[66]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[98]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[130]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[162]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[194]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[226]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[18]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[50]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[82]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[114]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[146]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[178]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[210]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[242]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[3]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[35]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[67]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[99]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[131]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[163]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[195]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[227]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[19]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[51]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[83]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[115]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[147]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[179]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[211]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[243]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[4]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[36]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[68]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[100]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[132]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[164]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[196]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[228]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[20]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[52]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[84]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[116]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[148]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[180]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[212]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[244]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[5]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[37]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[69]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[101]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[133]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[165]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[197]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[229]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[21]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[53]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[85]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[117]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[149]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[181]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[213]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[245]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[6]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[38]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[70]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[102]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[134]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[166]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[198]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[230]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[22]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[54]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[86]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[118]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[150]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[182]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[214]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[246]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[7]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[39]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[71]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[103]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[135]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[167]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[199]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[231]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[23]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[55]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[87]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[119]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[151]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[183]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[215]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[247]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[8]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[40]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[72]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[104]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[136]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[168]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[200]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[232]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[24]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[56]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[88]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[120]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[152]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[184]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[216]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[248]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[9]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[41]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[73]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[105]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[137]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[169]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[201]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[233]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[25]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[57]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[89]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[121]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[153]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[185]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[217]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[249]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[10]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[42]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[74]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[106]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[138]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[170]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[202]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[234]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[26]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[58]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[90]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[122]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[154]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[186]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[218]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[250]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[11]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[43]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[75]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[107]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[139]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[171]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[203]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[235]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[27]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[59]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[91]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[123]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[155]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[187]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[219]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[251]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[12]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[44]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[76]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[108]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[140]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[172]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[204]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[236]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[28]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[60]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[92]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[124]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[156]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[188]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[220]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[252]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[13]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[45]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[77]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[109]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[141]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[173]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[205]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[237]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[29]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[61]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[93]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[125]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[157]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[189]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[221]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[253]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[14]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[46]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[78]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[110]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[142]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[174]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[206]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[238]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[30]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[62]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[94]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[126]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[158]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[190]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[222]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[254]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[15]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[47]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[79]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[111]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[143]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[175]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[207]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[239]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[31]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[63]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[95]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[127]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[159]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[191]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[223]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[255]));
+      __syncthreads();
+      pad_temp_shared[((int)threadIdx.x)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) - 7)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 49)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 42)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 98)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 91)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 147)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 140)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 196)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 189)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 245)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 238)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 294)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 287)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 343)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 336)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 392)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 385)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 441)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 434)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 490)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 483)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 539)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 532)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 588)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 581)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 637)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 630)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 686)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 679)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 735)] = (((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 728)] : 0.000000e+00f);
+      kernel_shared[(((int)threadIdx.x) * 4)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 1)];
+      kernel_shared[((((int)threadIdx.x) * 4) + 1)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 10)];
+      kernel_shared[((((int)threadIdx.x) * 4) + 2)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 19)];
+      kernel_shared[((((int)threadIdx.x) * 4) + 3)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 28)];
+      if (((int)threadIdx.x) &lt; 15) {
+        kernel_shared[((((int)threadIdx.x) * 4) + 196)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 4) &amp; 15) * 9)) + (ry_outer_outer * 3)) + 1)];
+        kernel_shared[((((int)threadIdx.x) * 4) + 197)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 5) &amp; 15) * 9)) + (ry_outer_outer * 3)) + 1)];
+        kernel_shared[((((int)threadIdx.x) * 4) + 198)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 6) &amp; 15) * 9)) + (ry_outer_outer * 3)) + 1)];
+        kernel_shared[((((int)threadIdx.x) * 4) + 199)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 7) &amp; 15) * 9)) + (ry_outer_outer * 3)) + 1)];
+      }
+      __syncthreads();
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[0]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[32]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[64]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[96]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[128]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[160]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[192]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[224]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[16]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[48]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[80]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[112]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[144]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[176]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[208]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[240]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[1]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[33]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[65]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[97]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[129]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[161]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[193]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[225]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[17]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[49]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[81]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[113]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[145]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[177]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[209]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[241]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[2]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[34]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[66]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[98]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[130]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[162]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[194]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[226]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[18]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[50]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[82]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[114]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[146]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[178]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[210]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[242]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[3]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[35]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[67]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[99]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[131]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[163]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[195]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[227]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[19]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[51]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[83]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[115]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[147]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[179]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[211]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[243]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[4]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[36]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[68]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[100]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[132]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[164]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[196]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[228]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[20]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[52]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[84]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[116]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[148]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[180]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[212]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[244]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[5]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[37]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[69]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[101]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[133]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[165]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[197]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[229]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[21]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[53]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[85]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[117]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[149]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[181]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[213]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[245]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[6]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[38]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[70]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[102]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[134]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[166]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[198]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[230]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[22]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[54]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[86]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[118]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[150]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[182]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[214]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[246]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[7]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[39]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[71]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[103]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[135]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[167]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[199]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[231]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[23]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[55]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[87]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[119]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[151]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[183]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[215]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[247]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[8]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[40]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[72]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[104]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[136]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[168]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[200]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[232]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[24]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[56]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[88]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[120]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[152]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[184]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[216]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[248]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[9]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[41]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[73]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[105]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[137]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[169]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[201]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[233]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[25]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[57]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[89]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[121]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[153]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[185]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[217]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[249]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[10]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[42]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[74]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[106]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[138]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[170]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[202]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[234]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[26]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[58]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[90]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[122]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[154]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[186]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[218]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[250]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[11]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[43]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[75]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[107]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[139]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[171]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[203]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[235]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[27]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[59]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[91]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[123]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[155]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[187]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[219]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[251]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[12]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[44]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[76]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[108]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[140]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[172]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[204]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[236]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[28]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[60]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[92]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[124]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[156]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[188]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[220]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[252]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[13]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[45]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[77]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[109]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[141]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[173]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[205]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[237]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[29]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[61]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[93]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[125]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[157]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[189]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[221]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[253]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[14]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[46]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[78]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[110]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[142]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[174]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[206]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[238]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[30]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[62]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[94]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[126]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[158]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[190]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[222]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[254]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[15]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[47]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[79]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[111]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[143]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[175]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[207]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[239]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[31]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[63]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[95]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[127]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[159]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[191]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[223]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[255]));
+      __syncthreads();
+      pad_temp_shared[((int)threadIdx.x)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) - 6)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 49)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 43)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 98)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 92)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 147)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 141)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 196)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 190)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 245)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 239)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 294)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 288)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 343)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 337)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 392)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 386)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 441)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 435)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 490)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 484)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 539)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 533)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 588)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 582)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 637)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 631)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 686)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 680)] : 0.000000e+00f);
+      pad_temp_shared[(((int)threadIdx.x) + 735)] = ((((1 &lt;= ((((int)threadIdx.x) / 7) + ry_outer_outer)) &amp;&amp; (((((int)threadIdx.x) / 7) + ry_outer_outer) &lt; 8)) &amp;&amp; ((((int)threadIdx.x) % 7) &lt; 6)) ? data[((((rc_outer_outer * 784) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + 729)] : 0.000000e+00f);
+      kernel_shared[(((int)threadIdx.x) * 4)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 2)];
+      kernel_shared[((((int)threadIdx.x) * 4) + 1)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 11)];
+      kernel_shared[((((int)threadIdx.x) * 4) + 2)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 20)];
+      kernel_shared[((((int)threadIdx.x) * 4) + 3)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((int)threadIdx.x) &amp; 3) * 36)) + (ry_outer_outer * 3)) + 29)];
+      if (((int)threadIdx.x) &lt; 15) {
+        kernel_shared[((((int)threadIdx.x) * 4) + 196)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 4) &amp; 15) * 9)) + (ry_outer_outer * 3)) + 2)];
+        kernel_shared[((((int)threadIdx.x) * 4) + 197)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 5) &amp; 15) * 9)) + (ry_outer_outer * 3)) + 2)];
+        kernel_shared[((((int)threadIdx.x) * 4) + 198)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 6) &amp; 15) * 9)) + (ry_outer_outer * 3)) + 2)];
+        kernel_shared[((((int)threadIdx.x) * 4) + 199)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) &gt;&gt; 2) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) * 4) + 7) &amp; 15) * 9)) + (ry_outer_outer * 3)) + 2)];
+      }
+      __syncthreads();
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[0]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[32]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[64]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[96]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[128]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[160]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[192]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[224]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[16]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[48]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[80]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[112]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[144]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[176]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[208]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((int)threadIdx.x)] * kernel_shared[240]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[1]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[33]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[65]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[97]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[129]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[161]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[193]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[225]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[17]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[49]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[81]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[113]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[145]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[177]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[209]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 49)] * kernel_shared[241]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[2]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[34]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[66]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[98]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[130]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[162]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[194]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[226]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[18]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[50]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[82]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[114]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[146]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[178]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[210]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 98)] * kernel_shared[242]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[3]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[35]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[67]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[99]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[131]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[163]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[195]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[227]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[19]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[51]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[83]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[115]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[147]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[179]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[211]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 147)] * kernel_shared[243]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[4]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[36]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[68]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[100]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[132]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[164]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[196]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[228]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[20]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[52]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[84]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[116]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[148]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[180]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[212]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 196)] * kernel_shared[244]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[5]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[37]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[69]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[101]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[133]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[165]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[197]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[229]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[21]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[53]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[85]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[117]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[149]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[181]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[213]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 245)] * kernel_shared[245]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[6]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[38]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[70]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[102]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[134]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[166]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[198]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[230]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[22]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[54]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[86]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[118]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[150]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[182]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[214]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 294)] * kernel_shared[246]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[7]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[39]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[71]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[103]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[135]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[167]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[199]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[231]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[23]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[55]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[87]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[119]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[151]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[183]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[215]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 343)] * kernel_shared[247]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[8]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[40]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[72]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[104]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[136]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[168]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[200]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[232]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[24]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[56]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[88]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[120]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[152]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[184]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[216]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 392)] * kernel_shared[248]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[9]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[41]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[73]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[105]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[137]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[169]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[201]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[233]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[25]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[57]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[89]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[121]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[153]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[185]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[217]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 441)] * kernel_shared[249]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[10]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[42]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[74]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[106]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[138]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[170]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[202]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[234]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[26]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[58]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[90]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[122]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[154]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[186]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[218]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 490)] * kernel_shared[250]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[11]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[43]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[75]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[107]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[139]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[171]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[203]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[235]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[27]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[59]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[91]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[123]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[155]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[187]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[219]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 539)] * kernel_shared[251]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[12]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[44]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[76]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[108]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[140]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[172]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[204]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[236]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[28]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[60]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[92]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[124]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[156]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[188]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[220]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 588)] * kernel_shared[252]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[13]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[45]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[77]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[109]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[141]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[173]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[205]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[237]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[29]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[61]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[93]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[125]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[157]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[189]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[221]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 637)] * kernel_shared[253]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[14]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[46]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[78]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[110]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[142]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[174]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[206]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[238]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[30]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[62]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[94]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[126]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[158]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[190]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[222]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 686)] * kernel_shared[254]));
+      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[15]));
+      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[47]));
+      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[79]));
+      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[111]));
+      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[143]));
+      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[175]));
+      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[207]));
+      conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[239]));
+      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[31]));
+      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[63]));
+      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[95]));
+      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[127]));
+      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[159]));
+      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[191]));
+      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[223]));
+      conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((int)threadIdx.x) + 735)] * kernel_shared[255]));
     }
-    __syncthreads();
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[0]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[9]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[36]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[45]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[72]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[81]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7))] * kernel_shared[108]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[117]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[1]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[10]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[37]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[46]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[73]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[82]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[109]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 82)] * kernel_shared[118]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[2]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[11]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[38]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[47]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[74]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[83]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[110]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 83)] * kernel_shared[119]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[3]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[12]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[39]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[48]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[75]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[84]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[111]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[120]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[4]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[13]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[40]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[49]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[76]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[85]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 10)] * kernel_shared[112]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 91)] * kernel_shared[121]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[5]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[14]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[41]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[50]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[77]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[86]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 11)] * kernel_shared[113]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 92)] * kernel_shared[122]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[6]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[15]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[42]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[51]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[78]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[87]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[114]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[123]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[7]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[16]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[43]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[52]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[79]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[88]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 19)] * kernel_shared[115]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 100)] * kernel_shared[124]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[8]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[17]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[44]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[53]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[80]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[89]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 20)] * kernel_shared[116]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 101)] * kernel_shared[125]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[18]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[27]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[54]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[63]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[90]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[99]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[126]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[135]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[19]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[28]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[55]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[64]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[91]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[100]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 163)] * kernel_shared[127]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 244)] * kernel_shared[136]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[20]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[29]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[56]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[65]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[92]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[101]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 164)] * kernel_shared[128]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 245)] * kernel_shared[137]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[21]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[30]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[57]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[66]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[93]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[102]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[129]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[138]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[22]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[31]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[58]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[67]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[94]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[103]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 172)] * kernel_shared[130]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 253)] * kernel_shared[139]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[23]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[32]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[59]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[68]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[95]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[104]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 173)] * kernel_shared[131]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 254)] * kernel_shared[140]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[24]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[33]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[60]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[69]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[96]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[105]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[132]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[141]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[25]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[34]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[61]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[70]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[97]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[106]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 181)] * kernel_shared[133]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 262)] * kernel_shared[142]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[26]));
-    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[35]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[62]));
-    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[71]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[98]));
-    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[107]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 182)] * kernel_shared[134]));
-    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + (((int)threadIdx.x) % 7)) + 263)] * kernel_shared[143]));
   }
-  for (int i1_inner = 0; i1_inner &lt; 4; ++i1_inner) {
-    compute[(((((int)blockIdx.x) * 196) + (i1_inner * 49)) + ((int)threadIdx.x))] = max((conv2d_nchw[i1_inner] + bias[((((int)blockIdx.x) * 4) + i1_inner)]), 0.000000e+00f);
+  for (int i1_inner = 0; i1_inner &lt; 2; ++i1_inner) {
+    compute[(((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x))] = max((conv2d_nchw[i1_inner] + bias[((((int)blockIdx.x) * 16) + i1_inner)]), 0.000000e+00f);
+    compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 98)] = max((conv2d_nchw[(i1_inner + 2)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 2)]), 0.000000e+00f);
+    compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 196)] = max((conv2d_nchw[(i1_inner + 4)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 4)]), 0.000000e+00f);
+    compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 294)] = max((conv2d_nchw[(i1_inner + 6)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 6)]), 0.000000e+00f);
+    compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 392)] = max((conv2d_nchw[(i1_inner + 8)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 8)]), 0.000000e+00f);
+    compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 490)] = max((conv2d_nchw[(i1_inner + 10)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 10)]), 0.000000e+00f);
+    compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 588)] = max((conv2d_nchw[(i1_inner + 12)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 12)]), 0.000000e+00f);
+    compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 686)] = max((conv2d_nchw[(i1_inner + 14)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 14)]), 0.000000e+00f);
   }
 }
 </pre></div>
@@ -1033,10 +2507,9 @@ In the example below we resume the status and do more 5 trials.</p>
 /venv/apache-tvm-py3.7/lib/python3.7/site-packages/xgboost/training.py:17: UserWarning: Old style callback is deprecated.  See: https://xgboost.readthedocs.io/en/latest/python/callbacks.html
   warnings.warn(f&#39;Old style callback is deprecated.  See: {link}&#39;, UserWarning)
 Get devices for measurement successfully!
-.T
 </pre></div>
 </div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 5 minutes  52.239 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 5 minutes  43.754 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 da62ad4398..43e614a31c 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
@@ -915,7 +915,7 @@ so we can read the log file and load the best schedules.</p>
 Evaluate inference time cost...
 Execution time summary:
  mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)
-   8.2691       8.2682       8.2786       8.2605       0.0074
+   8.2044       8.2046       8.2096       8.1990       0.0043
 </pre></div>
 </div>
 </div>
@@ -937,7 +937,7 @@ 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  4.373 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  4.752 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 afeacb21a0..c60a1acb15 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
@@ -934,7 +934,7 @@ 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)
-  765.1622     765.7585     766.9518     762.7763      1.7560
+  763.1550     763.9276     764.1947     761.3427      1.2861
 </pre></div>
 </div>
 </div>
@@ -956,7 +956,7 @@ 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  34.879 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  34.813 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 7dd844db9b..df4119901d 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
@@ -632,30 +632,76 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
              placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [128, 512], []),
              compute: Buffer(compute_2: Pointer(float32), float32, [128, 512], [])}
   buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute} {
-  for (i0.outer.i1.outer.fused: int32, 0, 128) &quot;parallel&quot; {
-    allocate(compute_3: Pointer(global float32), float32, [512]), storage_scope = global {
-      for (i.outer.inner: int32, 0, 8) {
-        for (i.inner.init: int32, 0, 4) {
-          for (j.init: int32, 0, 16) {
-            compute_4: Buffer(compute_3, float32, [512], [])[(((i.outer.inner*64) + (i.inner.init*16)) + j.init)] = 0f32
+  for (i0.outer.i1.outer.fused: int32, 0, 256) &quot;parallel&quot; {
+    allocate(compute_3: Pointer(global float32), float32, [256]), storage_scope = global {
+      for (nb_j.inner: int32, 0, 2) {
+        for (i.inner.init: int32, 0, 8) {
+          let cse_var_1: int32 = ((i.inner.init*32) + (nb_j.inner*16))
+           {
+            compute_4: Buffer(compute_3, float32, [256], [])[cse_var_1] = 0f32
+            compute_4[(cse_var_1 + 1)] = 0f32
+            compute_4[(cse_var_1 + 2)] = 0f32
+            compute_4[(cse_var_1 + 3)] = 0f32
+            compute_4[(cse_var_1 + 4)] = 0f32
+            compute_4[(cse_var_1 + 5)] = 0f32
+            compute_4[(cse_var_1 + 6)] = 0f32
+            compute_4[(cse_var_1 + 7)] = 0f32
+            compute_4[(cse_var_1 + 8)] = 0f32
+            compute_4[(cse_var_1 + 9)] = 0f32
+            compute_4[(cse_var_1 + 10)] = 0f32
+            compute_4[(cse_var_1 + 11)] = 0f32
+            compute_4[(cse_var_1 + 12)] = 0f32
+            compute_4[(cse_var_1 + 13)] = 0f32
+            compute_4[(cse_var_1 + 14)] = 0f32
+            compute_4[(cse_var_1 + 15)] = 0f32
           }
         }
-        for (elem_idx: int32, 0, let cse_var_1: int32 = floormod(i0.outer.i1.outer.fused, 32) in (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(cse_var_1 + 1)] - placeholder_15[cse_var_1])) {
-          for (i.inner: int32, 0, 4) {
-            for (j: int32, 0, 16) {
-              let cse_var_2: int32 = floormod(i0.outer.i1.outer.fused, 32)
-              if @tir.likely((elem_idx &lt; (placeholder_15[(cse_var_2 + 1)] - placeholder_15[cse_var_2])), dtype=bool) {
-                let cse_var_3: int32 = (((i.outer.inner*64) + (i.inner*16)) + j)
-                compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[(((placeholder_15[cse_var_2]*16) + (elem_idx*16)) + j)]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[((((floordiv(i0.outer.i1.outer.fused, 32)*8192) + (i.outer.inner*1024)) + (i.inner*256)) + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[cse_var_2] + elem_idx)])], 0f32)))
-              }
+        for (elem_idx: int32, 0, let cse_var_2: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner) in (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(cse_var_2 + 1)] - placeholder_15[cse_var_2])) {
+          for (i.inner: int32, 0, 8) {
+            let cse_var_21: int32 = (elem_idx*16)
+            let cse_var_20: int32 = ((i.inner*32) + (nb_j.inner*16))
+            let cse_var_19: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner)
+            let cse_var_18: int32 = ((floordiv(i0.outer.i1.outer.fused, 16)*2048) + (i.inner*256))
+            let cse_var_17: int32 = (cse_var_20 + 9)
+            let cse_var_16: int32 = (cse_var_20 + 8)
+            let cse_var_15: int32 = (cse_var_20 + 7)
+            let cse_var_14: int32 = (cse_var_20 + 6)
+            let cse_var_13: int32 = (cse_var_20 + 5)
+            let cse_var_12: int32 = (cse_var_20 + 4)
+            let cse_var_11: int32 = (cse_var_20 + 3)
+            let cse_var_10: int32 = (cse_var_20 + 2)
+            let cse_var_9: int32 = (cse_var_20 + 15)
+            let cse_var_8: int32 = (cse_var_20 + 14)
+            let cse_var_7: int32 = (cse_var_20 + 13)
+            let cse_var_6: int32 = (cse_var_20 + 12)
+            let cse_var_5: int32 = (cse_var_20 + 11)
+            let cse_var_4: int32 = (cse_var_20 + 10)
+            let cse_var_3: int32 = (cse_var_20 + 1)
+             {
+              compute_4[cse_var_20] = (compute_4[cse_var_20] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[((placeholder_15[cse_var_19]*16) + cse_var_21)]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[(cse_var_18 + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 1)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_10] = (compute_4[cse_var_10] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 2)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_11] = (compute_4[cse_var_11] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 3)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_12] = (compute_4[cse_var_12] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 4)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_13] = (compute_4[cse_var_13] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 5)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_14] = (compute_4[cse_var_14] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 6)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_15] = (compute_4[cse_var_15] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 7)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_16] = (compute_4[cse_var_16] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 8)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_17] = (compute_4[cse_var_17] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 9)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_4] = (compute_4[cse_var_4] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 10)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_5] = (compute_4[cse_var_5] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 11)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_6] = (compute_4[cse_var_6] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 12)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_7] = (compute_4[cse_var_7] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 13)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_8] = (compute_4[cse_var_8] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 14)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
+              compute_4[cse_var_9] = (compute_4[cse_var_9] + (placeholder_16[(((placeholder_15[cse_var_19]*16) + cse_var_21) + 15)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_19] + elem_idx)])], 0f32)))
             }
           }
         }
       }
-      for (i0.inner: int32, 0, 32) {
-        for (i1.inner: int32, 0, 16) {
-          let cse_var_4: int32 = ((((floordiv(i0.outer.i1.outer.fused, 32)*16384) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 32)*16)) + i1.inner)
-          compute_5: Buffer(compute_2, float32, [65536], [])[cse_var_4] = max((compute_4[((i0.inner*16) + i1.inner)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[cse_var_4]), 0f32)
+      for (i0.inner: int32, 0, 8) {
+        for (i1.inner: int32, 0, 32) {
+          let cse_var_22: int32 = ((((floordiv(i0.outer.i1.outer.fused, 16)*4096) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 16)*32)) + i1.inner)
+          compute_5: Buffer(compute_2, float32, [65536], [])[cse_var_22] = max((compute_4[((i0.inner*32) + i1.inner)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[cse_var_22]), 0f32)
         }
       }
     }
@@ -694,7 +740,7 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
 <span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 1.477 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 1.906 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 055eb9b536..25273eb709 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>00:31.477</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
+<p><strong>00:35.755</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 84%" />
@@ -349,11 +349,11 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="tune_conv2d_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-conv2d-cuda-py"><span class="std std-ref">Tuning High Performance Convolution on NVIDIA GPUs</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_conv2d_cuda.py</span></code>)</p></td>
-<td><p>00:31.441</p></td>
+<td><p>00:35.720</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>00:00.021</p></td>
+<td><p>00:00.019</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="tune_relay_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-cuda-py"><span class="std std-ref">Auto-tuning a Convolutional Network for NVIDIA GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_cuda.py</span></code>)</p></td>
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 56c32a9a51..8fe20e60e2 100644
--- a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
+++ b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
@@ -567,7 +567,8 @@ for this template</p>
 waiting for device...
 device available
 Get devices for measurement successfully!
-No: 1   GFLOPS: 0.00/0.00       result: Traceback (most recent call last):
+No: 1   GFLOPS: 2.32/2.32       result: MeasureResult(costs=(0.099728897,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.013670921325684, timestamp=1668620461.9634821) [(&#39;tile_f&#39;, [-1, 8, 8, 4]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 1, 2]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,1781024
+No: 2   GFLOPS: 0.00/2.32       result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -689,8 +690,9 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 2, 32, 2]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 2, 64]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 0)],None,3647031
-No: 2   GFLOPS: 0.00/0.00       result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 16, 1, 8]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 2, 128]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 1)],None,10440020
+No: 3   GFLOPS: 1.98/2.32       result: MeasureResult(costs=(0.11707182299999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.8137314319610596, timestamp=1668620464.7798214)        [(&#39;tile_f&#39;, [-1, 128, 1, 1]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 4, 2]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,2561687
+No: 4   GFLOPS: 0.00/2.32       result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -812,9 +814,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 1, 8, 64]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 2, 256]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 1)],None,5803149
-No: 3   GFLOPS: 228.85/228.85   result: MeasureResult(costs=(0.0010115684646464648,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.749254941940308, timestamp=1668618422.8077989)       [(&#39;tile_f&#39;, [-1, 4, 2, 1]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 8, 2]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 1)],None,7984692
-No: 4   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 64, 1, 2]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 256, 2]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,3355501
+No: 5   GFLOPS: 0.00/2.32       result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -936,8 +937,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 1, 128, 4]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 16, 32]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 0)],None,350595
-No: 5   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 2, 2, 32]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 8, 16]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,3036631
+No: 6   GFLOPS: 0.00/2.32       result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -1059,9 +1060,9 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 1, 16, 8]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 2, 64]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 1)],None,9650018
-No: 6   GFLOPS: 126.37/228.85   result: MeasureResult(costs=(0.001831866727272727,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.34965181350708, timestamp=1668618425.2790372) [(&#39;tile_f&#39;, [-1, 1, 4, 4]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 8, 2]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,2177455
-No: 7   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 8, 8, 8]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 512, 1]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 1)],None,7778477
+No: 7   GFLOPS: 10.83/10.83     result: MeasureResult(costs=(0.021368804833333335,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6941542625427246, timestamp=1668620468.6158936)       [(&#39;tile_f&#39;, [-1, 32, 8, 1]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 1, 2]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,3326432
+No: 8   GFLOPS: 0.00/10.83      result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -1183,8 +1184,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 16, 4, 4]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 8, 16]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 1)],None,7101279
-No: 8   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 256, 2, 1]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 16, 8]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 1)],None,6304998
+No: 9   GFLOPS: 0.00/10.83      result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -1306,8 +1307,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 2, 128, 2]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 32, 2]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 0)],None,5089138
-No: 9   GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 64, 4, 1]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 256, 1]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 0)],None,609645
+No: 10  GFLOPS: 0.00/10.83      result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -1429,8 +1430,10 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 32, 4, 4]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 8, 64]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 0)],None,943480
-No: 10  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 4, 4, 4]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 16, 4]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 1)],None,9569677
+No: 11  GFLOPS: 160.31/160.31   result: MeasureResult(costs=(0.0014441192545454544,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.183053970336914, timestamp=1668620471.017138)        [(&#39;tile_f&#39;, [-1, 2, 4, 2]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 1, 16]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 1)],None,9412773
+No: 12  GFLOPS: 15.38/160.31    result: MeasureResult(costs=(0.015050043285714285,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.1918892860412598, timestamp=1668620471.784751)        [(&#39;tile_f&#39;, [-1, 8, 16, 4]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 2, 1]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,2520449
+No: 13  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -1552,8 +1555,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 4, 64, 2]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 1, 32]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 1)],None,7111376
-No: 11  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 8, 2, 8]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 16, 1]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 0)],None,4276066
+No: 14  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -1675,8 +1678,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 128, 2, 2]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 64, 2]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 1)],None,8769711
-No: 12  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 8, 4, 8]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 16, 32]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,3255932
+No: 15  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -1798,8 +1801,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 1, 128, 4]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 32, 8]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 1)],None,8437575
-No: 13  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 8, 32, 2]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 2, 16]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 0)],None,3803233
+No: 16  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -1921,8 +1924,9 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 4, 2, 2]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 4, 32]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 1)],None,6925666
-No: 14  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 8, 1, 32]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 16, 32]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 1)],None,5578288
+No: 17  GFLOPS: 36.23/160.31    result: MeasureResult(costs=(0.006389735090909091,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.769231081008911, timestamp=1668620476.9480226)        [(&#39;tile_f&#39;, [-1, 1, 4, 4]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 2, 16]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 0)],None,4576115
+No: 18  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -2044,8 +2048,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 1, 1, 512]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 4, 64]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,2489299
-No: 15  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 4, 1, 8]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 2, 64]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 1)],None,7713998
+No: 19  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -2167,8 +2171,8 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 256, 2, 1]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 16, 32]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,2867938
-No: 16  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 2, 2, 32]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 2, 32]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 1)],None,8664231
+No: 20  GFLOPS: 0.00/160.31     result: Traceback (most recent call last):
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
     func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
@@ -2290,133 +2294,7 @@ Traceback (most recent call last):
   File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
   File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
     raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 1, 256, 2]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 1, 64]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 1)],None,9065199
-No: 17  GFLOPS: 53.80/228.85    result: MeasureResult(costs=(0.00430309384,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4398438930511475, timestamp=1668618429.2793698)      [(&#39;tile_f&#39;, [-1, 1, 64, 8]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 2, 1]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 3]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 1)],None,6780343
-No: 18  GFLOPS: 0.00/228.85     result: Traceback (most recent call last):
-  File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 588, in __call__
-    func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
-  File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 540, in _build_func_common
-    func = build(s, args, target_host=task.target_host, runtime=runtime)
-  File &quot;/workspace/python/tvm/driver/build_module.py&quot;, line 227, in build
-    input_mod = lower(inputs, args, name=name, binds=binds)
-  File &quot;/workspace/python/tvm/driver/build_module.py&quot;, line 134, in lower
-    return ffi.lower_schedule(inp, args, name, binds, simple_mode)
-  File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
-  File &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 276, in tvm._ffi._cy3.core.FuncCall
-  File &quot;tvm/_ffi/_cython/./base.pxi&quot;, 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:1731
-  20: unpack_call&lt;tvm::IRModule, 5, tvm::&lt;lambda(tvm::te::Schedule, const tvm::runtime::Array&lt;tvm::runtime::ObjectRef&gt;&amp;, const tvm::runtime::String&amp;, const tvm::runtime::Map&lt;tvm::te::Tensor, tvm::tir::Buffer&gt;&amp;, bool)&gt; &gt;
-        at ../include/tvm/runtime/packed_func.h:1671
-  19: run&lt;&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  18: run&lt;tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  17: run&lt;tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  16: run&lt;tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  15: run&lt;tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  14: run&lt;tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1646
-  13: operator()
-        at ../src/driver/driver_api.cc:388
-  12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array&lt;tvm::runtime::ObjectRef, void&gt; const&amp;, std::__cxx11::basic_string&lt;char, std::char_traits&lt;char&gt;, std::allocator&lt;char&gt; &gt; const&amp;, std::unordered_map&lt;tvm::te::Tensor, tvm::tir::Buffer, std::hash&lt;tvm::te::Tensor&gt;, std::equal_to&lt;tvm::te::Tensor&gt;, std::allocator&lt;std::pair&lt;tvm::te::Tensor const, tvm::tir::Buffer&gt; &gt; &gt; const&amp;, tvm::GlobalVarSupply, bool)
-        at ../src/driver/driver_api.cc:374
-  11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array&lt;tvm::transform::Pass, void&gt;)
-        at ../src/driver/driver_api.cc:269
-  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&amp;) const
-        at ../src/ir/transform.cc:274
-  8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&amp;) const
-        at ../src/ir/transform.cc:453
-  7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&amp;) const
-        at ../src/ir/transform.cc:274
-  6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&amp;) const
-        at ../src/tir/ir/transform.cc:100
-  5: tvm::runtime::TypedPackedFunc&lt;tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)&gt;::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
-        at ../include/tvm/runtime/packed_func.h:1750
-  4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher&lt;tvm::tir::PrimFunc&gt;::run&lt;tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext&gt;(tvm::runtime::PackedFunc const&amp;, tvm::tir::PrimFunc&amp;&amp;, tvm::IRModule&amp;&amp;, tvm::transform::PassContext&amp;&amp;)
-        at ../include/tvm/runtime/packed_func.h:1694
-  3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()&lt;tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext&gt;(tvm::tir::PrimFunc&amp;&amp;, tvm::IRModule&amp;&amp;, tvm::transform::PassContext&amp;&amp;) const
-        at ../include/tvm/runtime/packed_func.h:1618
-  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 &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
-  File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
-    raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-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:1731
-  20: unpack_call&lt;tvm::IRModule, 5, tvm::&lt;lambda(tvm::te::Schedule, const tvm::runtime::Array&lt;tvm::runtime::ObjectRef&gt;&amp;, const tvm::runtime::String&amp;, const tvm::runtime::Map&lt;tvm::te::Tensor, tvm::tir::Buffer&gt;&amp;, bool)&gt; &gt;
-        at ../include/tvm/runtime/packed_func.h:1671
-  19: run&lt;&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  18: run&lt;tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  17: run&lt;tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  16: run&lt;tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  15: run&lt;tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1631
-  14: run&lt;tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_&gt;
-        at ../include/tvm/runtime/packed_func.h:1646
-  13: operator()
-        at ../src/driver/driver_api.cc:388
-  12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array&lt;tvm::runtime::ObjectRef, void&gt; const&amp;, std::__cxx11::basic_string&lt;char, std::char_traits&lt;char&gt;, std::allocator&lt;char&gt; &gt; const&amp;, std::unordered_map&lt;tvm::te::Tensor, tvm::tir::Buffer, std::hash&lt;tvm::te::Tensor&gt;, std::equal_to&lt;tvm::te::Tensor&gt;, std::allocator&lt;std::pair&lt;tvm::te::Tensor const, tvm::tir::Buffer&gt; &gt; &gt; const&amp;, tvm::GlobalVarSupply, bool)
-        at ../src/driver/driver_api.cc:374
-  11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array&lt;tvm::transform::Pass, void&gt;)
-        at ../src/driver/driver_api.cc:269
-  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&amp;) const
-        at ../src/ir/transform.cc:274
-  8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&amp;) const
-        at ../src/ir/transform.cc:453
-  7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&amp;) const
-        at ../src/ir/transform.cc:274
-  6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&amp;) const
-        at ../src/tir/ir/transform.cc:100
-  5: tvm::runtime::TypedPackedFunc&lt;tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)&gt;::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
-        at ../include/tvm/runtime/packed_func.h:1750
-  4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher&lt;tvm::tir::PrimFunc&gt;::run&lt;tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext&gt;(tvm::runtime::PackedFunc const&amp;, tvm::tir::PrimFunc&amp;&amp;, tvm::IRModule&amp;&amp;, tvm::transform::PassContext&amp;&amp;)
-        at ../include/tvm/runtime/packed_func.h:1694
-  3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()&lt;tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext&gt;(tvm::tir::PrimFunc&amp;&amp;, tvm::IRModule&amp;&amp;, tvm::transform::PassContext&amp;&amp;) const
-        at ../include/tvm/runtime/packed_func.h:1618
-  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 &quot;tvm/_ffi/_cython/./packed_func.pxi&quot;, line 56, in tvm._ffi._cy3.core.tvm_callback
-  File &quot;/workspace/python/tvm/autotvm/measure/measure_methods.py&quot;, line 871, in verify_pass
-    raise InstantiationError(&quot;Skipped because of invalid gpu kernel&quot;)
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 4, 4, 16]), (&#39;tile_y&#39;, [-1, 1, 1, 7]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 64, 2]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 0)],None,639717
-No: 19  GFLOPS: 37.57/228.85    result: MeasureResult(costs=(0.006161984117647059,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.3533060550689697, timestamp=1668618430.00153) [(&#39;tile_f&#39;, [-1, 32, 2, 1]), (&#39;tile_y&#39;, [-1, 1, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 7, 1]), (&#39;tile_rc&#39;, [-1, 4, 2]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 1, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 1)],None,5658415
-No: 20  GFLOPS: 2.70/228.85     result: MeasureResult(costs=(0.085883934,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.988248109817505, timestamp=1668618431.4606593) [(&#39;tile_f&#39;, [-1, 32, 2, 1]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 7]), (&#39;tile_rc&#39;, [-1, 1, 2]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 0)],None,2554875
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [(&#39;tile_f&#39;, [-1, 1, 256, 1]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 8, 64]), (&#39;tile_ry&#39;, [-1, 3, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 0), (&#39;unroll_explicit&#39;, 0)],None,943632
 </pre></div>
 </div>
 <p>Finally we can inspect the best config from log file, check correctness,
@@ -2455,9 +2333,9 @@ 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:
-[(&#39;tile_f&#39;, [-1, 4, 2, 1]), (&#39;tile_y&#39;, [-1, 1, 7, 1]), (&#39;tile_x&#39;, [-1, 7, 1, 1]), (&#39;tile_rc&#39;, [-1, 8, 2]), (&#39;tile_ry&#39;, [-1, 1, 3]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 512), (&#39;unroll_explicit&#39;, 1)],None,7984692
+[(&#39;tile_f&#39;, [-1, 2, 4, 2]), (&#39;tile_y&#39;, [-1, 7, 1, 1]), (&#39;tile_x&#39;, [-1, 1, 1, 1]), (&#39;tile_rc&#39;, [-1, 1, 16]), (&#39;tile_ry&#39;, [-1, 1, 1]), (&#39;tile_rx&#39;, [-1, 3, 1]), (&#39;auto_unroll_max_step&#39;, 1500), (&#39;unroll_explicit&#39;, 1)],None,9412773
 Finish loading 20 records
-Time cost of this operator: 0.000809
+Time cost of this operator: 0.001880
 </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 9420514b7c..bef57d13b5 100644
--- a/docs/how_to/work_with_microtvm/micro_autotune.html
+++ b/docs/how_to/work_with_microtvm/micro_autotune.html
@@ -596,10 +596,10 @@ 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  313.9     98.712   (1, 2, 10, 10, 3)  2       1        [313.9]
-tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       3.131     0.985    (1, 6, 10, 10)     1       1        [3.131]
-tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.964     0.303    (1, 1, 10, 10, 3)  1       1        [0.964]
-Total_time                                    -                                             317.996   -        -                  -       -        -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  309.4     98.728   (1, 2, 10, 10, 3)  2       1        [309.4]
+tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       3.023     0.965    (1, 6, 10, 10)     1       1        [3.023]
+tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.962     0.307    (1, 1, 10, 10, 3)  1       1        [0.962]
+Total_time                                    -                                             313.386   -        -                  -       -        -
 </pre></div>
 </div>
 </div>
@@ -650,10 +650,10 @@ Total_time                                    -
 <div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>########## Build with Autotuning ##########
 Node Name                                     Ops                                           Time(us)  Time(%)  Shape              Inputs  Outputs  Measurements(us)
 ---------                                     ---                                           --------  -------  -----              ------  -------  ----------------
-tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  102.6     97.476   (1, 6, 10, 10, 1)  2       1        [102.6]
-tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       1.799     1.709    (1, 6, 10, 10)     1       1        [1.799]
-tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.858     0.815    (1, 3, 10, 10, 1)  1       1        [0.858]
-Total_time                                    -                                             105.257   -        -                  -       -        -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  101.8     97.407   (1, 6, 10, 10, 1)  2       1        [101.8]
+tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       1.756     1.68     (1, 6, 10, 10)     1       1        [1.756]
+tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.955     0.913    (1, 1, 10, 10, 3)  1       1        [0.955]
+Total_time                                    -                                             104.51    -        -                  -       -        -
 </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 3f71482e11..245a0bfd4f 100644
--- a/docs/how_to/work_with_microtvm/micro_pytorch.html
+++ b/docs/how_to/work_with_microtvm/micro_pytorch.html
@@ -440,7 +440,8 @@ download a cat image and preprocess it to use as the model input.</p>
 Downloading: &quot;https://download.pytorch.org/models/quantized/mobilenet_v2_qnnpack_37f702c5.pth&quot; to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2_qnnpack_37f702c5.pth
 
   0%|          | 0.00/3.42M [00:00&lt;?, ?B/s]
-100%|##########| 3.42M/3.42M [00:00&lt;00:00, 75.6MB/s]
+ 76%|#######6  | 2.61M/3.42M [00:00&lt;00:00, 27.3MB/s]
+100%|##########| 3.42M/3.42M [00:00&lt;00:00, 27.9MB/s]
 /workspace/python/tvm/relay/frontend/pytorch_utils.py:47: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.
   return LooseVersion(torch_ver) &gt; 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.
@@ -564,7 +565,7 @@ 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  4.241 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  5.341 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 1a27ffe458..d63332177b 100644
--- a/docs/how_to/work_with_microtvm/micro_train.html
+++ b/docs/how_to/work_with_microtvm/micro_train.html
@@ -530,7 +530,7 @@ 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">&quot;</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>&#39;/tmp/tmpbz6hkicy/images/random&#39;
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>&#39;/tmp/tmpe0nb0235/images/random&#39;
 </pre></div>
 </div>
 </div>
@@ -590,8 +590,8 @@ 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">&quot;off&quot;</span><span class="p">)</span>
 </pre></div>
 </div>
-<img src="../../_images/sphx_glr_micro_train_001.png" srcset="../../_images/sphx_glr_micro_train_001.png" alt="[0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0]" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>/tmp/tmpbz6hkicy/images/target contains 8144 images
-/tmp/tmpbz6hkicy/images/random contains 5000 images
+<img src="../../_images/sphx_glr_micro_train_001.png" srcset="../../_images/sphx_glr_micro_train_001.png" alt="[0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0]" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>/tmp/tmpe0nb0235/images/target contains 8144 images
+/tmp/tmpe0nb0235/images/random contains 5000 images
 </pre></div>
 </div>
 </div>
@@ -703,13 +703,13 @@ 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: 0.2374 - accuracy: 0.9209 - val_loss: 0.1806 - val_accuracy: 0.9388 - 47s/epoch - 143ms/step
+328/328 - 47s - loss: 0.2109 - accuracy: 0.9260 - val_loss: 0.1198 - val_accuracy: 0.9569 - 47s/epoch - 144ms/step
 Epoch 2/3
-328/328 - 44s - loss: 0.1106 - accuracy: 0.9570 - val_loss: 0.0969 - val_accuracy: 0.9641 - 44s/epoch - 133ms/step
+328/328 - 44s - loss: 0.1042 - accuracy: 0.9624 - val_loss: 0.1374 - val_accuracy: 0.9498 - 44s/epoch - 133ms/step
 Epoch 3/3
-328/328 - 43s - loss: 0.0756 - accuracy: 0.9724 - val_loss: 0.0981 - val_accuracy: 0.9690 - 43s/epoch - 132ms/step
+328/328 - 44s - loss: 0.0650 - accuracy: 0.9738 - val_loss: 0.1059 - val_accuracy: 0.9660 - 44s/epoch - 133ms/step
 
-&lt;keras.callbacks.History object at 0x7f5cc00bee90&gt;
+&lt;keras.callbacks.History object at 0x7f34dd15a3d0&gt;
 </pre></div>
 </div>
 </div>
@@ -971,7 +971,7 @@ as intended.</p>
 <p>From here, we could modify the model to read live images from the camera - we have another
 Arduino tutorial for how to do that <a class="reference external" href="https://github.com/guberti/tvm-arduino-demos/tree/master/examples/person_detection">on GitHub</a>. Alternatively, we could also
 <a class="reference external" href="https://tvm.apache.org/docs/how_to/work_with_microtvm/micro_autotune.html">use TVM’s autotuning capabilities</a> to dramatically improve the model’s performance.</p>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 4 minutes  47.987 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 4 minutes  42.129 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 3a8020d28a..8880d52b3f 100644
--- a/docs/how_to/work_with_microtvm/sg_execution_times.html
+++ b/docs/how_to/work_with_microtvm/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-work-with-microtvm-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>06:54.261</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
+<p><strong>06:51.708</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 83%" />
@@ -349,23 +349,23 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="micro_train.html#sphx-glr-how-to-work-with-microtvm-micro-train-py"><span class="std std-ref">Training Vision Models for microTVM on Arduino</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_train.py</span></code>)</p></td>
-<td><p>04:47.987</p></td>
+<td><p>04:42.129</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="micro_pytorch.html#sphx-glr-how-to-work-with-microtvm-micro-pytorch-py"><span class="std std-ref">microTVM PyTorch Tutorial</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_pytorch.py</span></code>)</p></td>
-<td><p>01:04.241</p></td>
+<td><p>01:05.341</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="micro_autotune.html#sphx-glr-how-to-work-with-microtvm-micro-autotune-py"><span class="std std-ref">Autotuning with microTVM</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_autotune.py</span></code>)</p></td>
-<td><p>00:49.982</p></td>
+<td><p>00:51.638</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="micro_aot.html#sphx-glr-how-to-work-with-microtvm-micro-aot-py"><span class="std std-ref">microTVM Host-Driven AoT</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_aot.py</span></code>)</p></td>
-<td><p>00:08.264</p></td>
+<td><p>00:08.628</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="micro_tflite.html#sphx-glr-how-to-work-with-microtvm-micro-tflite-py"><span class="std std-ref">microTVM with TFLite Models</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_tflite.py</span></code>)</p></td>
-<td><p>00:03.784</p></td>
+<td><p>00:03.969</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="micro_reference_vm.html#sphx-glr-how-to-work-with-microtvm-micro-reference-vm-py"><span class="std std-ref">microTVM Reference Virtual Machines</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_reference_vm.py</span></code>)</p></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 e11bb722c8..ab4df7f4d4 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>00:43.314</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
+<p><strong>00:44.784</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>00:31.790</p></td>
+<td><p>00:32.784</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>00:10.146</p></td>
+<td><p>00:10.272</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="build_gcn.html#sphx-glr-how-to-work-with-relay-build-gcn-py"><span class="std std-ref">Building a Graph Convolutional Network</span></a> (<code class="docutils literal notranslate"><span class="pre">build_gcn.py</span></code>)</p></td>
-<td><p>00:01.372</p></td>
+<td><p>00:01.721</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 c388b9c6cb..4d5b751c11 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">&quot;tir.exp&quot;</span><span class="p">,</span> <span class="n">target</span><span class="o">=</span><span class="s2">&quot;cuda&quot;</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>&lt;function my_cuda_math_rule at 0x7f5cc13808c0&gt;
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>&lt;function my_cuda_math_rule at 0x7f330c705e60&gt;
 </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 2aab511e27..04c12e1015 100644
--- a/docs/how_to/work_with_schedules/sg_execution_times.html
+++ b/docs/how_to/work_with_schedules/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-work-with-schedules-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:07.373</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
+<p><strong>00:06.448</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 83%" />
@@ -349,35 +349,35 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="intrin_math.html#sphx-glr-how-to-work-with-schedules-intrin-math-py"><span class="std std-ref">Intrinsics and Math Functions</span></a> (<code class="docutils literal notranslate"><span class="pre">intrin_math.py</span></code>)</p></td>
-<td><p>00:05.026</p></td>
+<td><p>00:03.974</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>00:01.043</p></td>
+<td><p>00:01.086</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>00:00.555</p></td>
+<td><p>00:00.594</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>00:00.535</p></td>
+<td><p>00:00.577</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>00:00.116</p></td>
+<td><p>00:00.118</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="schedule_primitives.html#sphx-glr-how-to-work-with-schedules-schedule-primitives-py"><span class="std std-ref">Schedule Primitives in TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">schedule_primitives.py</span></code>)</p></td>
-<td><p>00:00.049</p></td>
+<td><p>00:00.050</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="tedd.html#sphx-glr-how-to-work-with-schedules-tedd-py"><span class="std std-ref">Use Tensor Expression Debug Display (TEDD) for Visualization</span></a> (<code class="docutils literal notranslate"><span class="pre">tedd.py</span></code>)</p></td>
-<td><p>00:00.029</p></td>
+<td><p>00:00.030</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="tuple_inputs.html#sphx-glr-how-to-work-with-schedules-tuple-inputs-py"><span class="std std-ref">Compute and Reduce with Tuple Inputs</span></a> (<code class="docutils literal notranslate"><span class="pre">tuple_inputs.py</span></code>)</p></td>
-<td><p>00:00.019</p></td>
+<td><p>00:00.020</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 </tbody>
diff --git a/docs/how_to/work_with_schedules/tensorize.html b/docs/how_to/work_with_schedules/tensorize.html
index a1a1d3fbe0..f28b0c0f34 100644
--- a/docs/how_to/work_with_schedules/tensorize.html
+++ b/docs/how_to/work_with_schedules/tensorize.html
@@ -586,7 +586,7 @@ The importing needs to happen before the tensorized GEMV being executed.</p>
              B: Buffer(B_2: Pointer(float32), float32, [512, 64], []),
              C: Buffer(C_2: Pointer(float32), float32, [1024, 512], [])}
   buffer_map = {A_1: A, B_1: B, C_1: C} {
-  attr [IterVar(i: int32, (nullptr), &quot;DataPar&quot;, &quot;&quot;)] &quot;pragma_import_llvm&quot; = &quot;; ModuleID = &#39;/tmp/tmp0b8qkn1u/input0.cc&#39;\nsource_filename = \&quot;/tmp/tmp0b8qkn1u/input0.cc\&quot;\ntarget datalayout = \&quot;e-m:e-i64:64-f80:128-n8:16:32:64-S128\&quot;\ntarget triple = \&quot;x86_64-pc-linux-gnu\&quot;\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n  %7 = allo [...]
+  attr [IterVar(i: int32, (nullptr), &quot;DataPar&quot;, &quot;&quot;)] &quot;pragma_import_llvm&quot; = &quot;; ModuleID = &#39;/tmp/tmp8j_p51zl/input0.cc&#39;\nsource_filename = \&quot;/tmp/tmp8j_p51zl/input0.cc\&quot;\ntarget datalayout = \&quot;e-m:e-i64:64-f80:128-n8:16:32:64-S128\&quot;\ntarget triple = \&quot;x86_64-pc-linux-gnu\&quot;\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n  %7 = allo [...]
   for (i, 0, 1024) {
     for (j.outer: int32, 0, 32) {
       @tir.call_extern(&quot;gemv_update&quot;, @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), C_2, ((i*512) + (j.outer*16)), 16, 2, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), A_2, (i*64), 64, 1, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), B_2, (j.outer*1024), 1024, 1, dtype=handle), 16, 64, 64, dtype=int32)
diff --git a/docs/install/nnpack.html b/docs/install/nnpack.html
index 23d2181e9d..1ef28de467 100644
--- a/docs/install/nnpack.html
+++ b/docs/install/nnpack.html
@@ -229,17 +229,7 @@
               <p class="caption" role="heading"><span class="caption-text">Getting Started</span></p>
 <ul class="current">
 <li class="toctree-l1 current"><a class="reference internal" href="index.html">Installing TVM</a><ul class="current">
-<li class="toctree-l2 current"><a class="reference internal" href="from_source.html">Install from Source</a><ul class="current">
-<li class="toctree-l3"><a class="reference internal" href="from_source.html#developers-get-source-from-github">Developers: Get Source from Github</a></li>
-<li class="toctree-l3"><a class="reference internal" href="from_source.html#build-the-shared-library">Build the Shared Library</a></li>
-<li class="toctree-l3"><a class="reference internal" href="from_source.html#python-package-installation">Python Package Installation</a></li>
-<li class="toctree-l3 current"><a class="reference internal" href="from_source.html#install-contrib-libraries">Install Contrib Libraries</a><ul class="current">
-<li class="toctree-l4 current"><a class="current reference internal" href="#">NNPACK Contrib Installation</a></li>
-</ul>
-</li>
-<li class="toctree-l3"><a class="reference internal" href="from_source.html#enable-c-tests">Enable C++ Tests</a></li>
-</ul>
-</li>
+<li class="toctree-l2"><a class="reference internal" href="from_source.html">Install from Source</a></li>
 <li class="toctree-l2"><a class="reference internal" href="docker.html">Docker Images</a></li>
 <li class="toctree-l2 current"><a class="current reference internal" href="#">NNPACK Contrib Installation</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="#conditions">Conditions</a></li>
diff --git a/docs/reference/api/python/auto_scheduler.html b/docs/reference/api/python/auto_scheduler.html
index 86a37e1231..99337088f8 100644
--- a/docs/reference/api/python/auto_scheduler.html
+++ b/docs/reference/api/python/auto_scheduler.html
@@ -1615,7 +1615,7 @@ history states as starting point to perform Evolutionary Search).</p></li>
 
 <dl class="py class">
 <dt class="sig sig-object py" id="tvm.auto_scheduler.SketchPolicy">
-<em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">tvm.auto_scheduler.</span></span><span class="sig-name descname"><span class="pre">SketchPolicy</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">task</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">program_cost_model</span></span><span class="o"><span class="pre">=</span></span><span class="defau [...]
+<em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">tvm.auto_scheduler.</span></span><span class="sig-name descname"><span class="pre">SketchPolicy</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">task</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">program_cost_model</span></span><span class="o"><span class="pre">=</span></span><span class="defau [...]
 <dd><p>The search policy that searches in a hierarchical search space defined by sketches.
 The policy randomly samples programs from the space defined by sketches and use evolutionary
 search to fine-tune them.</p>
@@ -1899,7 +1899,7 @@ Candidates:
 
 <dl class="py function">
 <dt class="sig sig-object py" id="tvm.auto_scheduler.auto_schedule">
-<span class="sig-prename descclassname"><span class="pre">tvm.auto_scheduler.</span></span><span class="sig-name descname"><span class="pre">auto_schedule</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">task</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">search_policy</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em clas [...]
+<span class="sig-prename descclassname"><span class="pre">tvm.auto_scheduler.</span></span><span class="sig-name descname"><span class="pre">auto_schedule</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">task</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">search_policy</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em clas [...]
 <dd><p>THIS API IS DEPRECATED.</p>
 <p>Run auto scheduling search for a task.</p>
 <dl class="field-list simple">
diff --git a/docs/reference/api/typedoc/classes/bytestreamreader.html b/docs/reference/api/typedoc/classes/bytestreamreader.html
index 7c304c0740..6f2e1db736 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/44ed06ac9/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L42">rpc_server.ts:42</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L63">rpc_server.ts:63</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L49">rpc_server.ts:49</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L57">rpc_server.ts:57</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 7fff7c266f..932a65b88f 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/44ed06ac9/web/src/memory.ts#L223">memory.ts:223</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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">&lt;</span><a href="../interfaces/disposable.html" class="tsd-signature-type">Disposable</a><span class="tsd-signature-symbol">&gt;</span><span class="tsd-signature-symbol"> = []</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/memory.ts#L208">memory.ts:208</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L312">memory.ts:312</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L284">memory.ts:284</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L388">memory.ts:388</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L376">memory.ts:376</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L267">memory.ts:267</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L243">memory.ts:243</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L321">memory.ts:321</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L252">memory.ts:252</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L359">memory.ts:359</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L342">memory.ts:342</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L350">memory.ts:350</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L326">memory.ts:326</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L363">memory.ts:363</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L346">memory.ts:346</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L334">memory.ts:334</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 60b86565f7..70053c7b20 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/44ed06ac9/web/src/runtime.ts#L262">runtime.ts:262</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L260">runtime.ts:260</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L258">runtime.ts:258</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L262">runtime.ts:262</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L279">runtime.ts:279</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L270">runtime.ts:270</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 f0d545637b..0228a9dc09 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/44ed06ac9/web/src/runtime.ts#L202">runtime.ts:202</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L200">runtime.ts:200</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L198">runtime.ts:198</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L223">runtime.ts:223</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L230">runtime.ts:230</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 a2322feef4..4efbf7782f 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/44ed06ac9/web/src/environment.ts#L86">environment.ts:86</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/environment.ts#L70">environment.ts:70</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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"> =&gt; </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/44ed06ac9/web/src/environment.ts#L69">environment.ts:69</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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">&lt;</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">&gt;</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/44ed06ac9/web/src/environment.ts#L78">environment.ts:78</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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">&lt;</span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">&gt;</span><span class="tsd-signature-symbol"> = []</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/environment.ts#L84">environment.ts:84</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/environment.ts#L105">environment.ts:105</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 90d2b9759e..8917a8decb 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/44ed06ac9/web/src/runtime.ts#L49">runtime.ts:49</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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">&lt;</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">&gt;</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L46">runtime.ts:46</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L45">runtime.ts:45</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L44">runtime.ts:44</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L47">runtime.ts:47</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L76">runtime.ts:76</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L66">runtime.ts:66</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L84">runtime.ts:84</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L95">runtime.ts:95</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L72">runtime.ts:72</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 5762841b2c..f3d2808f09 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/44ed06ac9/web/src/runtime.ts#L583">runtime.ts:583</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L579">runtime.ts:579</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L654">runtime.ts:654</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L597">runtime.ts:597</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L631">runtime.ts:631</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L644">runtime.ts:644</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L621">runtime.ts:621</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L609">runtime.ts:609</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 8023824a54..57589bd1f0 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/44ed06ac9/web/src/runtime.ts#L692">runtime.ts:692</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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">&lt;</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">&gt;</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L684">runtime.ts:684</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L683">runtime.ts:683</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L932">runtime.ts:932</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L994">runtime.ts:994</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L924">runtime.ts:924</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L732">runtime.ts:732</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L952">runtime.ts:952</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L816">runtime.ts:816</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L1033">runtime.ts:1033</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L846">runtime.ts:846</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L750">runtime.ts:750</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L1013">runtime.ts:1013</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L789">runtime.ts:789</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L914">runtime.ts:914</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L1145">runtime.ts:1145</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L740">runtime.ts:740</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L868">runtime.ts:868</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L857">runtime.ts:857</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L940">runtime.ts:940</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 b1f7857cb5..ae08e57e27 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/44ed06ac9/web/src/memory.ts#L40">memory.ts:40</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L32">memory.ts:32</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L33">memory.ts:33</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L154">memory.ts:154</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L90">memory.ts:90</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L97">memory.ts:97</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L74">memory.ts:74</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L81">memory.ts:81</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L104">memory.ts:104</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L132">memory.ts:132</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L145">memory.ts:145</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L60">memory.ts:60</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L67">memory.ts:67</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L53">memory.ts:53</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L114">memory.ts:114</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L124">memory.ts:124</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/memory.ts#L175">memory.ts:175</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 9474ac1e19..8a1ac06b6e 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/44ed06ac9/web/src/runtime.ts#L504">runtime.ts:504</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L502">runtime.ts:502</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L516">runtime.ts:516</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L530">runtime.ts:530</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L561">runtime.ts:561</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 f076e0af7c..da03326591 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/44ed06ac9/web/src/runtime.ts#L304">runtime.ts:304</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L297">runtime.ts:297</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L293">runtime.ts:293</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L289">runtime.ts:289</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L291">runtime.ts:291</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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">&lt;</span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">&gt;</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L295">runtime.ts:295</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L370">runtime.ts:370</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L414">runtime.ts:414</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L355">runtime.ts:355</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L474">runtime.ts:474</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L443">runtime.ts:443</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 af88505a4b..fd601e8faf 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/44ed06ac9/web/src/runtime.ts#L158">runtime.ts:158</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L157">runtime.ts:157</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L165">runtime.ts:165</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 02d681f2f1..46269058be 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/44ed06ac9/web/src/rpc_server.ts#L92">rpc_server.ts:92</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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"> =&gt; </span><span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol">&lt;</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/44ed06ac9/web/src/rpc_server.ts#L82">rpc_server.ts:82</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L78">rpc_server.ts:78</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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"> =&gt; </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/44ed06ac9/web/src/rpc_server.ts#L81">rpc_server.ts:81</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L79">rpc_server.ts:79</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L80">rpc_server.ts:80</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/rpc_server.ts#L77">rpc_server.ts:77</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 333171887e..7b05c6b17c 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/44ed06ac9/web/src/runtime.ts#L145">runtime.ts:145</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L145">runtime.ts:145</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/runtime.ts#L143">runtime.ts:143</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 245f826434..e8712071db 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/44ed06ac9/web/src/webgpu.ts#L57">webgpu.ts:57</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/webgpu.ts#L50">webgpu.ts:50</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/webgpu.ts#L51">webgpu.ts:51</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/webgpu.ts#L84">webgpu.ts:84</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/webgpu.ts#L170">webgpu.ts:170</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/webgpu.ts#L170">webgpu.ts:170</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/44ed06ac9/web/src/webgpu.ts#L67">webgpu.ts:67</a></li>
+									<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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 8d582641e6..7eb4380646 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/44ed06ac9/web/src/ctypes.ts#L220">ctypes.ts:220</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L216">ctypes.ts:216</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L214">ctypes.ts:214</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L218">ctypes.ts:218</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L226">ctypes.ts:226</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L221">ctypes.ts:221</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L219">ctypes.ts:219</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L223">ctypes.ts:223</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L227">ctypes.ts:227</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L222">ctypes.ts:222</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L228">ctypes.ts:228</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L217">ctypes.ts:217</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L224">ctypes.ts:224</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/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/44ed06ac9/web/src/ctypes.ts#L225">ctypes.ts:225</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L225">ctypes.ts:225</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -246,7 +246,7 @@
 					<div class="tsd-signature tsd-kind-icon">UInt<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 1</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L215">ctypes.ts:215</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L215">ctypes.ts:215</a></li>
 						</ul>
 					</aside>
 				</section>
diff --git a/docs/reference/api/typedoc/enums/aynccallbackcode.html b/docs/reference/api/typedoc/enums/aynccallbackcode.html
index b79f365c8b..088afaba00 100644
--- a/docs/reference/api/typedoc/enums/aynccallbackcode.html
+++ b/docs/reference/api/typedoc/enums/aynccallbackcode.html
@@ -93,7 +93,7 @@
 					<div class="tsd-signature tsd-kind-icon">k<wbr>Exception<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 5</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L676">runtime.ts:676</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/runtime.ts#L676">runtime.ts:676</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -103,7 +103,7 @@
 					<div class="tsd-signature tsd-kind-icon">k<wbr>Return<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 4</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L675">runtime.ts:675</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/runtime.ts#L675">runtime.ts:675</a></li>
 						</ul>
 					</aside>
 				</section>
diff --git a/docs/reference/api/typedoc/enums/dldatatypecode.html b/docs/reference/api/typedoc/enums/dldatatypecode.html
index 63af544e06..4694291e2f 100644
--- a/docs/reference/api/typedoc/enums/dldatatypecode.html
+++ b/docs/reference/api/typedoc/enums/dldatatypecode.html
@@ -95,7 +95,7 @@
 					<div class="tsd-signature tsd-kind-icon">Float<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 2</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L242">runtime.ts:242</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/runtime.ts#L242">runtime.ts:242</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -105,7 +105,7 @@
 					<div class="tsd-signature tsd-kind-icon">Int<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 0</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L240">runtime.ts:240</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/runtime.ts#L240">runtime.ts:240</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -115,7 +115,7 @@
 					<div class="tsd-signature tsd-kind-icon">Opaque<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 3</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L243">runtime.ts:243</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/runtime.ts#L243">runtime.ts:243</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -125,7 +125,7 @@
 					<div class="tsd-signature tsd-kind-icon">UInt<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 1</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/runtime.ts#L241">runtime.ts:241</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/runtime.ts#L241">runtime.ts:241</a></li>
 						</ul>
 					</aside>
 				</section>
diff --git a/docs/reference/api/typedoc/enums/rpcserverstate.html b/docs/reference/api/typedoc/enums/rpcserverstate.html
index 1f55842ba6..3c9533760a 100644
--- a/docs/reference/api/typedoc/enums/rpcserverstate.html
+++ b/docs/reference/api/typedoc/enums/rpcserverstate.html
@@ -90,7 +90,7 @@
 					<div class="tsd-signature tsd-kind-icon">Init<wbr>Header<span class="tsd-signature-symbol">:</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/rpc_server.ts#L27">rpc_server.ts:27</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/rpc_server.ts#L27">rpc_server.ts:27</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -100,7 +100,7 @@
 					<div class="tsd-signature tsd-kind-icon">Init<wbr>Header<wbr>Key<span class="tsd-signature-symbol">:</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/rpc_server.ts#L28">rpc_server.ts:28</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/rpc_server.ts#L28">rpc_server.ts:28</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -110,7 +110,7 @@
 					<div class="tsd-signature tsd-kind-icon">Init<wbr>Server<span class="tsd-signature-symbol">:</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/rpc_server.ts#L29">rpc_server.ts:29</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/rpc_server.ts#L29">rpc_server.ts:29</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -120,7 +120,7 @@
 					<div class="tsd-signature tsd-kind-icon">Receive<wbr>Packet<wbr>Body<span class="tsd-signature-symbol">:</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/rpc_server.ts#L32">rpc_server.ts:32</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/rpc_server.ts#L32">rpc_server.ts:32</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -130,7 +130,7 @@
 					<div class="tsd-signature tsd-kind-icon">Receive<wbr>Packet<wbr>Header<span class="tsd-signature-symbol">:</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/rpc_server.ts#L31">rpc_server.ts:31</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/rpc_server.ts#L31">rpc_server.ts:31</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -140,7 +140,7 @@
 					<div class="tsd-signature tsd-kind-icon">Wait<wbr>For<wbr>Callback<span class="tsd-signature-symbol">:</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/rpc_server.ts#L30">rpc_server.ts:30</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/rpc_server.ts#L30">rpc_server.ts:30</a></li>
 						</ul>
 					</aside>
 				</section>
diff --git a/docs/reference/api/typedoc/enums/sizeof.html b/docs/reference/api/typedoc/enums/sizeof.html
index 335fa6ad3b..2dd5a43908 100644
--- a/docs/reference/api/typedoc/enums/sizeof.html
+++ b/docs/reference/api/typedoc/enums/sizeof.html
@@ -100,7 +100,7 @@
 					<div class="tsd-signature tsd-kind-icon">DLData<wbr>Type<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = I32</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L206">ctypes.ts:206</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L206">ctypes.ts:206</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -110,7 +110,7 @@
 					<div class="tsd-signature tsd-kind-icon">DLDevice<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = I32 + I32</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L207">ctypes.ts:207</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L207">ctypes.ts:207</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -120,7 +120,7 @@
 					<div class="tsd-signature tsd-kind-icon">F32<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/44ed06ac9/web/src/ctypes.ts#L203">ctypes.ts:203</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L203">ctypes.ts:203</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -130,7 +130,7 @@
 					<div class="tsd-signature tsd-kind-icon">F64<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/44ed06ac9/web/src/ctypes.ts#L204">ctypes.ts:204</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L204">ctypes.ts:204</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -140,7 +140,7 @@
 					<div class="tsd-signature tsd-kind-icon">I32<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/44ed06ac9/web/src/ctypes.ts#L201">ctypes.ts:201</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L201">ctypes.ts:201</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -150,7 +150,7 @@
 					<div class="tsd-signature tsd-kind-icon">I64<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/44ed06ac9/web/src/ctypes.ts#L202">ctypes.ts:202</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L202">ctypes.ts:202</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -160,7 +160,7 @@
 					<div class="tsd-signature tsd-kind-icon">TVMValue<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/44ed06ac9/web/src/ctypes.ts#L205">ctypes.ts:205</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L205">ctypes.ts:205</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -170,7 +170,7 @@
 					<div class="tsd-signature tsd-kind-icon">U16<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/44ed06ac9/web/src/ctypes.ts#L200">ctypes.ts:200</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L200">ctypes.ts:200</a></li>
 						</ul>
 					</aside>
 				</section>
@@ -180,7 +180,7 @@
 					<div class="tsd-signature tsd-kind-icon">U8<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 1</span></div>
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L199">ctypes.ts:199</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L199">ctypes.ts:199</a></li>
 						</ul>
 					</aside>
 				</section>
diff --git a/docs/reference/api/typedoc/index.html b/docs/reference/api/typedoc/index.html
index 429a388133..e23b580cb9 100644
--- a/docs/reference/api/typedoc/index.html
+++ b/docs/reference/api/typedoc/index.html
@@ -174,7 +174,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMArray<wbr>Alloc<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>shape<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, ndim<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">number</span>, dtypeCode<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">number</span>, dtypeBits<span class="tsd [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L112">ctypes.ts:112</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L112">ctypes.ts:112</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -238,7 +238,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMArray<wbr>Copy<wbr>From<wbr>Bytes<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>handle<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, data<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, nbytes<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">num [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L128">ctypes.ts:128</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L128">ctypes.ts:128</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -282,7 +282,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMArray<wbr>Copy<wbr>From<wbr>To<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>from<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, to<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, stream<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-sig [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L144">ctypes.ts:144</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L144">ctypes.ts:144</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -326,7 +326,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMArray<wbr>Copy<wbr>ToBytes<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>handle<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, data<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, nbytes<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">number</sp [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L136">ctypes.ts:136</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L136">ctypes.ts:136</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -370,7 +370,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMArray<wbr>Free<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>handle<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </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/44ed06ac9/web/src/ctypes.ts#L121">ctypes.ts:121</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L121">ctypes.ts:121</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -406,7 +406,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMBackend<wbr>PackedCFunc<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>argValues<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, argCodes<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, nargs<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">number< [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L160">ctypes.ts:160</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L160">ctypes.ts:160</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -458,7 +458,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMCFunc<wbr>Set<wbr>Return<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>ret<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, value<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, typeCode<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signa [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L77">ctypes.ts:77</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L77">ctypes.ts:77</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -506,7 +506,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMCb<wbr>Arg<wbr>ToReturn<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>value<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, code<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </span><span c [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L83">ctypes.ts:83</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L83">ctypes.ts:83</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -545,7 +545,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMFunc<wbr>Call<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>func<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, argValues<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, typeCode<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-t [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L67">ctypes.ts:67</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L67">ctypes.ts:67</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -601,7 +601,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMFunc<wbr>Free<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>func<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </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/44ed06ac9/web/src/ctypes.ts#L57">ctypes.ts:57</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L57">ctypes.ts:57</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -637,7 +637,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMFunc<wbr>Get<wbr>Global<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>name<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, out<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </span><span cla [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L100">ctypes.ts:100</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L100">ctypes.ts:100</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -676,7 +676,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMFunc<wbr>List<wbr>Global<wbr>Names<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>outSize<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, outArray<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&g [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L88">ctypes.ts:88</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L88">ctypes.ts:88</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -715,7 +715,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMFunc<wbr>Register<wbr>Global<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>name<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, f<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, override<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">number</spa [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L94">ctypes.ts:94</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L94">ctypes.ts:94</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -758,7 +758,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMGet<wbr>Last<wbr>Error<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </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/44ed06ac9/web/src/ctypes.ts#L34">ctypes.ts:34</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L34">ctypes.ts:34</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -788,7 +788,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMMod<wbr>Free<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>mod<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </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/44ed06ac9/web/src/ctypes.ts#L52">ctypes.ts:52</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L52">ctypes.ts:52</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -824,7 +824,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMMod<wbr>Get<wbr>Function<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>mod<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, funcName<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, queryImports<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">numbe [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L42">ctypes.ts:42</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L42">ctypes.ts:42</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -872,7 +872,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMMod<wbr>Import<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>mod<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a>, dep<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </span><span class="tsd-si [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L48">ctypes.ts:48</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L48">ctypes.ts:48</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -912,7 +912,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMSynchronize<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>deviceType<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">number</span>, deviceId<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">number</span>, stream<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signatur [...]
 					<aside class="tsd-sources">
 						<ul>
-							<li>Defined in <a href="https://github.com/apache/tvm/blob/44ed06ac9/web/src/ctypes.ts#L150">ctypes.ts:150</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L150">ctypes.ts:150</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -954,7 +954,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMWasm<wbr>Alloc<wbr>Space<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>size<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </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/44ed06ac9/web/src/ctypes.ts#L167">ctypes.ts:167</a></li>
+							<li>Defined in <a href="https://github.com/apache/tvm/blob/52739ef8c/web/src/ctypes.ts#L167">ctypes.ts:167</a></li>
 						</ul>
 					</aside>
 					<div class="tsd-comment tsd-typography">
@@ -990,7 +990,7 @@
 					<div class="tsd-signature tsd-kind-icon">FTVMWasm<wbr>Free<wbr>Space<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>ptr<span class="tsd-signature-symbol">: </span><a href="index.html#pointer" class="tsd-signature-type">Pointer</a><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> =&gt; </span><span class="tsd-signature-type">void</span></div>
... 1167 lines suppressed ...