You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by tq...@apache.org on 2022/09/14 17:52:40 UTC
[tvm-site] branch asf-site updated: deploying docs (apache/tvm@a40849342d250bd585e19434e4a2473fcf978bcb)
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 c670a53d5 deploying docs (apache/tvm@a40849342d250bd585e19434e4a2473fcf978bcb)
c670a53d5 is described below
commit c670a53d52b826b7ef1b8112f561629c3f9c67d2
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Wed Sep 14 17:52:33 2022 +0000
deploying docs (apache/tvm@a40849342d250bd585e19434e4a2473fcf978bcb)
---
.../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 | 4 +-
.../tune_network_arm.rst.txt | 298 ++---
.../tune_network_cuda.rst.txt | 332 +++---
.../tune_network_mali.rst.txt | 298 ++---
.../tune_network_x86.rst.txt | 410 +++----
.../tune_sparse_x86.rst.txt | 1183 +++++++-------------
.../tune_with_autotvm/sg_execution_times.rst.txt | 6 +-
.../tune_with_autotvm/tune_conv2d_cuda.rst.txt | 26 +-
.../work_with_microtvm/micro_autotune.rst.txt | 16 +-
.../how_to/work_with_microtvm/micro_tflite.rst.txt | 2 +-
.../how_to/work_with_microtvm/micro_train.rst.txt | 16 +-
.../work_with_microtvm/sg_execution_times.rst.txt | 10 +-
.../work_with_relay/sg_execution_times.rst.txt | 10 +-
.../how_to/work_with_schedules/intrin_math.rst.txt | 2 +-
.../work_with_schedules/sg_execution_times.rst.txt | 12 +-
.../how_to/work_with_schedules/tensorize.rst.txt | 2 +-
.../tutorials/autotvm/sg_execution_times.rst.txt | 4 +-
.../frontend/deploy_classification.rst.txt | 2 +-
.../tutorials/frontend/deploy_detection.rst.txt | 2 +-
.../tutorials/frontend/sg_execution_times.rst.txt | 6 +-
.../tutorials/optimize/sg_execution_times.rst.txt | 6 +-
.../topic/vta/tutorials/sg_execution_times.rst.txt | 6 +-
.../tutorial/auto_scheduler_matmul_x86.rst.txt | 6 +-
docs/_sources/tutorial/autotvm_matmul_x86.rst.txt | 20 +-
docs/_sources/tutorial/autotvm_relay_x86.rst.txt | 58 +-
.../tutorial/cross_compilation_and_rpc.rst.txt | 2 +-
docs/_sources/tutorial/intro_topi.rst.txt | 2 +-
docs/_sources/tutorial/sg_execution_times.rst.txt | 24 +-
.../tutorial/tensor_expr_get_started.rst.txt | 43 +-
docs/commit_hash | 2 +-
docs/how_to/compile_models/from_darknet.html | 2 +-
docs/how_to/compile_models/from_keras.html | 2 +-
docs/how_to/compile_models/from_mxnet.html | 2 +-
docs/how_to/compile_models/from_oneflow.html | 13 +-
docs/how_to/compile_models/from_pytorch.html | 5 +-
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 | 46 +-
docs/how_to/deploy_models/deploy_prequantized.html | 7 +-
.../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 | 38 +-
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 | 4 +-
.../tune_with_autoscheduler/tune_network_arm.html | 298 ++---
.../tune_with_autoscheduler/tune_network_cuda.html | 332 +++---
.../tune_with_autoscheduler/tune_network_mali.html | 298 ++---
.../tune_with_autoscheduler/tune_network_x86.html | 410 +++----
.../tune_with_autoscheduler/tune_sparse_x86.html | 1183 +++++++-------------
.../tune_with_autotvm/sg_execution_times.html | 6 +-
.../how_to/tune_with_autotvm/tune_conv2d_cuda.html | 26 +-
docs/how_to/work_with_microtvm/micro_autotune.html | 16 +-
docs/how_to/work_with_microtvm/micro_tflite.html | 2 +-
docs/how_to/work_with_microtvm/micro_train.html | 16 +-
.../work_with_microtvm/sg_execution_times.html | 10 +-
.../how_to/work_with_relay/sg_execution_times.html | 10 +-
docs/how_to/work_with_schedules/intrin_math.html | 2 +-
.../work_with_schedules/sg_execution_times.html | 12 +-
docs/how_to/work_with_schedules/tensorize.html | 2 +-
docs/reference/api/python/auto_scheduler.html | 4 +-
.../api/typedoc/classes/bytestreamreader.html | 12 +-
.../api/typedoc/classes/cachedcallstack.html | 34 +-
docs/reference/api/typedoc/classes/dldatatype.html | 12 +-
docs/reference/api/typedoc/classes/dldevice.html | 10 +-
.../reference/api/typedoc/classes/environment.html | 12 +-
docs/reference/api/typedoc/classes/ffilibrary.html | 20 +-
.../api/typedoc/classes/graphexecutor.html | 16 +-
docs/reference/api/typedoc/classes/instance.html | 40 +-
docs/reference/api/typedoc/classes/memory.html | 34 +-
docs/reference/api/typedoc/classes/module.html | 10 +-
docs/reference/api/typedoc/classes/ndarray.html | 22 +-
.../api/typedoc/classes/packedfunccell.html | 6 +-
docs/reference/api/typedoc/classes/rpcserver.html | 14 +-
docs/reference/api/typedoc/classes/scalar.html | 6 +-
.../api/typedoc/classes/webgpucontext.html | 12 +-
docs/reference/api/typedoc/enums/argtypecode.html | 30 +-
.../api/typedoc/enums/aynccallbackcode.html | 4 +-
.../api/typedoc/enums/dldatatypecode.html | 8 +-
.../api/typedoc/enums/rpcserverstate.html | 12 +-
docs/reference/api/typedoc/enums/sizeof.html | 18 +-
docs/reference/api/typedoc/index.html | 112 +-
.../api/typedoc/interfaces/disposable.html | 2 +-
.../api/typedoc/interfaces/functioninfo.html | 6 +-
.../api/typedoc/interfaces/libraryprovider.html | 4 +-
docs/searchindex.js | 2 +-
.../vta/tutorials/autotvm/sg_execution_times.html | 4 +-
.../tutorials/frontend/deploy_classification.html | 2 +-
.../vta/tutorials/frontend/deploy_detection.html | 2 +-
.../vta/tutorials/frontend/sg_execution_times.html | 6 +-
.../vta/tutorials/optimize/sg_execution_times.html | 6 +-
docs/topic/vta/tutorials/sg_execution_times.html | 6 +-
docs/tutorial/auto_scheduler_matmul_x86.html | 4 +-
docs/tutorial/autotvm_matmul_x86.html | 20 +-
docs/tutorial/autotvm_relay_x86.html | 262 ++---
docs/tutorial/cross_compilation_and_rpc.html | 2 +-
docs/tutorial/intro_topi.html | 2 +-
docs/tutorial/sg_execution_times.html | 28 +-
docs/tutorial/tensor_expr_get_started.html | 39 +-
129 files changed, 2962 insertions(+), 3695 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 4a09f39a9..d1ae372e6 100644
--- a/docs/_sources/how_to/compile_models/from_darknet.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_darknet.rst.txt
@@ -317,7 +317,7 @@ The process is no different from other examples.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 3.162 seconds)
+ **Total running time of the script:** ( 1 minutes 5.675 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 00780d31c..112869d65 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 945ms/step
+
1/1 [==============================] - ETA: 0s
1/1 [==============================] - 1s 963ms/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 4f417df8f..ae89c0272 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.zip616bf519-d84d-4695-a579-d4c9f206c94b from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+ Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip12d54117-6adf-40be-b795-3e3081d46c5e 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 8eeb3b3f0..fe3647504 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, 60.1MB/s]
29%|##9 | 12.1M/41.5M [00:00<00:00, 39.0MB/s]
45%|####5 | 18.8M/41.5M [00:00<00:00, 49.8MB/s]
58%|#####7 | 24.1M/41.5M [00:00<00:00, 48.0MB/s]
77%|#######7 | 32.0M/41.5M [00:00<00:00, 53.2MB/s]
96%|#########6| 40.0M/41.5M [00:00<00:00, 57.9MB/s]
100%|##########| 41.5M/41.5M [00:00<00:00, 54.8MB/s]
+
0%| | 0.00/41.5M [00:00<?, ?B/s]
19%|#9 | 7.99M/41.5M [00:00<00:00, 67.0MB/s]
39%|###8 | 16.0M/41.5M [00:00<00:00, 70.6MB/s]
58%|#####7 | 24.0M/41.5M [00:00<00:00, 47.4MB/s]
77%|#######7 | 32.0M/41.5M [00:00<00:00, 50.3MB/s]
90%|########9 | 37.2M/41.5M [00:00<00:00, 46.8MB/s]
100%|##########| 41.5M/41.5M [00:00<00:00, 47.8MB/s]
diff --git a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
index 48de54c54..41dfc209c 100644
--- a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
@@ -94,7 +94,7 @@ Load a pretrained PyTorch model
.. code-block:: none
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]
44%|####3 | 19.6M/44.7M [00:00<00:00, 205MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 242MB/s]
+
0%| | 0.00/44.7M [00:00<?, ?B/s]
39%|###9 | 17.6M/44.7M [00:00<00:00, 185MB/s]
93%|#########3| 41.6M/44.7M [00:00<00:00, 224MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 220MB/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 54c9aac18..3152a0fba 100644
--- a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
@@ -423,7 +423,7 @@ Run the corresponding model on tensorflow
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 6.729 seconds)
+ **Total running time of the script:** ( 1 minutes 9.617 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 92289f843..8fa2ef762 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
=================
-**05:08.365** total execution time for **how_to_compile_models** files:
+**05:21.075** total execution time for **how_to_compile_models** files:
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:06.729 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:09.617 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:03.162 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:05.675 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:39.177 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:41.350 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:28.318 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:28.858 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:25.761 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:27.428 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:24.882 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:24.557 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:21.841 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:23.199 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:19.904 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:20.335 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:16.112 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:17.210 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.480 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.846 | 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 102eed398..f850b8e35 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
@@ -441,7 +441,7 @@ Execute on TVM
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 15.9743 15.9829 16.0650 15.8152 0.0614
+ 16.1038 16.0971 16.1992 16.0414 0.0496
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 ca0db6438..7598f7f61 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
@@ -123,7 +123,7 @@ Load pre-trained maskrcnn from torchvision and do tracing
.. code-block:: none
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]
9%|9 | 16.0M/170M [00:00<00:00, 168MB/s]
22%|##2 | 37.9M/170M [00:00<00:00, 204MB/s]
35%|###5 | 60.0M/170M [00:00<00:00, 217MB/s]
49%|####8 | 82.5M/170M [00:00<00:00, 224MB/s]
61%|######1 | 104M/170M [00:00<00:00, 215MB/s]
73%|#######3 | 125M/170M [00:00<00:00, 188MB/s]
88%|########8 | 150M/170M [00:00<00:00, 211MB/s]
100%|##########| 170M/170M [00:00<00:00, 213MB/s]
+
0%| | 0.00/170M [00:00<?, ?B/s]
2%|2 | 3.57M/170M [00:00<00:04, 37.4MB/s]
4%|4 | 7.35M/170M [00:00<00:04, 38.7MB/s]
7%|6 | 11.0M/170M [00:00<00:04, 35.8MB/s]
10%|9 | 16.9M/170M [00:00<00:03, 45.3MB/s]
13%|#2 | 21.9M/170M [00:00<00:03, 47.6MB/s]
16%|#5 | 26.5M/170M [00:00<00:04, 35.8MB/s]
18%|#7 | 30.3M/170M [00:00<00:04, 36.3MB/s]
21%|## | 35.2M/170M [00:00<00:03, 40.4MB/s]
24%|##3 | 40.0M/170M [00:01<00:03, 42.9MB/s]
26%|##6 | 44.3M/170M [00:01<00:03, 41.0MB/s]
30%|##9 | 50.1M/170M [00:01<00:02, 46.6MB/s]
32%|###2 | 55.1M/170M [00:01<00:02, 48.3MB/s]
35%|###5 | 59.9M/170M [00:01<00:02, 47.5MB/s]
38%|###8 | 64.8M/170M [00:01<00:02, 48.6MB/s]
41%|####1 | 70.3M/170M [00:01<00:02, 51.3MB/s]
44%|####4 | 75.2M/170M [00:01<00:02, 48.8MB/s]
47%|####7 | 79.9M/170M [00:01<00:01, 47.3MB/
s]
50%|####9 | 84.5M/170M [00:02<00:02, 41.5MB/s]
52%|#####2 | 88.8M/170M [00:02<00:02, 42.3MB/s]
55%|#####4 | 93.0M/170M [00:02<00:01, 42.4MB/s]
57%|#####7 | 97.1M/170M [00:02<00:01, 39.3MB/s]
59%|#####9 | 101M/170M [00:02<00:01, 37.3MB/s]
62%|######1 | 105M/170M [00:02<00:01, 38.5MB/s]
64%|######4 | 109M/170M [00:02<00:01, 40.3MB/s]
67%|######7 | 114M/170M [00:02<00:01, 43.6MB/s]
71%|####### | 120M/170M [00:02<00:01, 47.6MB/s]
73%|#######3 | 124M/170M [00:03<00:01, 43.9MB/s]
76%|#######5 | 129M/170M [00:03<00:00, 43.2MB/s]
78%|#######8 | 133M/170M [00:03<00:00, 40.4MB/s]
81%|########1 | 138M/170M [00:03<00:00, 45.1MB/s]
84%|########4 | 143M/170M [00:03<00:00, 46.6MB/s]
88%|########7 | 149M/170M [00:03<00:00, 49.9MB/s]
91%|#########1| 155M/170M [00:03<00:00, 53.0MB/s]
94%|#########4| 160M/170M [00:03<00:00, 54.9MB/s]
97%|#########7| 166M/170M [00:03<00:00, 51.1MB/s]
100%|##########| 170M/170M [00:04<00:00, 44.2MB/s]
/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:3878: 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)
/usr/local/lib/python3.7/dist-packages/torchvision/models/detection/anchor_utils.py:127: 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').
@@ -295,7 +295,7 @@ Get boxes with score larger than 0.9
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 2 minutes 57.889 seconds)
+ **Total running time of the script:** ( 3 minutes 8.175 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 62fee551b..9e8281408 100644
--- a/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
@@ -232,7 +232,7 @@ training. Other models require a full post training calibration.
.. code-block:: none
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]
100%|##########| 13.6M/13.6M [00:00<00:00, 188MB/s]
+
0%| | 0.00/13.6M [00:00<?, ?B/s]
63%|######3 | 8.59M/13.6M [00:00<00:00, 90.0MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 108MB/s]
@@ -412,7 +412,7 @@ 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)
- 90.1986 90.0857 94.0330 89.9495 0.4338
+ 90.4280 90.3556 91.4413 90.1687 0.2261
@@ -461,7 +461,7 @@ TODO
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 10.116 seconds)
+ **Total running time of the script:** ( 1 minutes 11.984 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 4791554d8..b5a09733a 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
@@ -439,7 +439,7 @@ 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)
- 120.0377 119.9317 122.4276 119.3749 0.4780
+ 122.5429 122.5292 124.1027 121.9173 0.3705
@@ -476,7 +476,7 @@ 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:** ( 1 minutes 54.857 seconds)
+ **Total running time of the script:** ( 1 minutes 55.655 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 da44e330b..f8668bcde 100644
--- a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
@@ -255,7 +255,7 @@ We create a Relay VM to build and execute the model.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 20.366 seconds)
+ **Total running time of the script:** ( 1 minutes 24.558 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 61757334c..f1bc5479c 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
@@ -158,7 +158,7 @@ Convert and compile model for CPU.
data: None
input_sym_arg_type = in_param.infer_type()[0]
Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/ssd_512_resnet50_v1_voc-9c8b225a.zip...
-
0%| | 0/132723 [00:00<?, ?KB/s]
4%|4 | 5372/132723 [00:00<00:02, 53716.47KB/s]
10%|9 | 12665/132723 [00:00<00:01, 65008.43KB/s]
15%|#5 | 19948/132723 [00:00<00:01, 68577.37KB/s]
21%|## | 27267/132723 [00:00<00:01, 70396.61KB/s]
26%|##6 | 34604/132723 [00:00<00:01, 71465.66KB/s]
32%|###1 | 42043/132723 [00:00<00:01, 72448.79KB/s]
37%|###7 | 49437/132723 [00:00<00:01, 72932.56KB/s]
43%|####2 | 56833/132723 [00:00<00:01, 73255.51KB/s]
48%|####8 | 64174/132723 [00:00<00:00, 73300.65KB/s]
54%|#####3 | 71541/132723 [00:01<00:00, 73413.50KB/s]
59%|#####9 | 78883/132723 [00:01<00:00, 73381.35KB/s]
65%|######4 | 86244/132723 [00:01<00:00, 73450.00KB/s]
71%|####### | 93591/132723 [00:01<00:00, 73453.37KB/s]
76%|#######6 | 100964/132723 [00:01<00:00, 73528.83KB/s]
82%|########1 | 108344/132723 [00:01<00:00, 73604.01KB/s]
87%|########7
| 115705/132723 [00:01<00:00, 73480.31KB/s]
93%|#########2| 123091/132723 [00:01<00:00, 73592.93KB/s]
98%|#########8| 130451/132723 [00:01<00:00, 73492.71KB/s]
100%|##########| 132723/132723 [00:01<00:00, 72445.75KB/s]
+
0%| | 0/132723 [00:00<?, ?KB/s]
5%|4 | 6061/132723 [00:00<00:02, 60605.00KB/s]
11%|# | 14279/132723 [00:00<00:01, 73291.78KB/s]
16%|#6 | 21609/132723 [00:00<00:01, 65455.35KB/s]
23%|##2 | 29950/132723 [00:00<00:01, 72063.22KB/s]
29%|##8 | 38148/132723 [00:00<00:01, 75478.19KB/s]
35%|###5 | 46525/132723 [00:00<00:01, 78211.55KB/s]
41%|#### | 54408/132723 [00:00<00:01, 78091.50KB/s]
47%|####7 | 62829/132723 [00:00<00:00, 80004.30KB/s]
53%|#####3 | 70861/132723 [00:00<00:00, 79510.75KB/s]
60%|#####9 | 79182/132723 [00:01<00:00, 80636.24KB/s]
66%|######6 | 87601/132723 [00:01<00:00, 81712.50KB/s]
72%|#######2 | 96039/132723 [00:01<00:00, 82516.09KB/s]
79%|#######8 | 104432/132723 [00:01<00:00, 82939.84KB/s]
85%|########5 | 112850/132723 [00:01<00:00, 83312.31KB/s]
91%|#########1| 121186/132723 [00:01<00:00, 72318.24KB/s]
98%|########
#7| 129463/132723 [00:01<00:00, 75154.44KB/s]
100%|##########| 132723/132723 [00:01<00:00, 77096.56KB/s]
@@ -241,7 +241,7 @@ Display result
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 2 minutes 38.412 seconds)
+ **Total running time of the script:** ( 2 minutes 42.368 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 76554629f..69289b8a9 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
=================
-**11:16.656** total execution time for **how_to_deploy_models** files:
+**11:40.421** 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``) | 02:57.889 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:08.175 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 02:38.412 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 02:42.368 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 01:54.857 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 01:55.655 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:20.366 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:24.558 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:10.116 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:11.984 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:30.446 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:31.951 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:22.595 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:23.137 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:21.968 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:22.586 | 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 ff3bfbd8e..d1a008029 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
@@ -476,7 +476,7 @@ 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.zip2311a753-afdc-4829-87a6-73829e254cc7 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+ Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip670083a9-8dde-4f6e-bbc7-64ce068df60e 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 7d867b1ae..ef6a7e0d3 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
=================
-**00:41.558** total execution time for **how_to_extend_tvm** files:
+**00:41.775** 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``) | 00:38.390 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:38.553 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.213 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.245 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:00.947 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:00.969 | 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 59755926a..7bfaeab7c 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: 6729us [6729us] (45.91%; 45.91%)
- FoldScaleAxis: 7929us [5us] (54.09%; 54.09%)
- FoldConstant: 7923us [1676us] (54.06%; 99.93%)
- InferType: 6247us [6247us] (42.62%; 78.84%)
+ InferType: 6757us [6757us] (45.94%; 45.94%)
+ FoldScaleAxis: 7953us [6us] (54.06%; 54.06%)
+ FoldConstant: 7947us [1671us] (54.02%; 99.93%)
+ InferType: 6276us [6276us] (42.67%; 78.98%)
@@ -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: 6313us [6313us] (42.74%; 42.74%)
- FoldScaleAxis: 8459us [5us] (57.26%; 57.26%)
- FoldConstant: 8453us [1653us] (57.23%; 99.94%)
- InferType: 6800us [6800us] (46.03%; 80.44%)
+ InferType: 6362us [6362us] (44.65%; 44.65%)
+ FoldScaleAxis: 7885us [5us] (55.35%; 55.35%)
+ FoldConstant: 7880us [1682us] (55.31%; 99.94%)
+ InferType: 6198us [6198us] (43.51%; 78.66%)
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 3b60fca99..9b308671e 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: 54.210207 ms
+ Convolution: 37.319583 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 aac7dc5a7..8d4feef59 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
@@ -671,7 +671,7 @@ be able to run on our build server
.. code-block:: none
- conv2d with tensor core: 7.739008 ms
+ conv2d with tensor core: 12.897894 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 8031d04e0..39d4f009f 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: 0.018910
- Baseline: 3.256800
+ Numpy running time: 0.019168
+ Baseline: 3.438547
@@ -239,7 +239,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
.. code-block:: none
- Opt1: 0.310534
+ Opt1: 0.328304
@@ -342,7 +342,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
.. code-block:: none
- Opt2: 0.351062
+ Opt2: 0.350647
@@ -438,7 +438,7 @@ the access pattern for A matrix is more cache friendly.
.. code-block:: none
- Opt3: 0.113828
+ Opt3: 0.121355
@@ -563,7 +563,7 @@ flattening.
.. code-block:: none
- Opt4: 0.108118
+ Opt4: 0.109599
@@ -685,7 +685,7 @@ write to C when all the block results are ready.
.. code-block:: none
- Opt5: 0.111650
+ Opt5: 0.112299
@@ -810,7 +810,7 @@ Furthermore, we can also utilize multi-core processors to do the thread-level pa
.. code-block:: none
- Opt6: 0.146985
+ Opt6: 0.148837
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 a07538b3d..37d59fa6e 100644
--- a/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
@@ -5,12 +5,12 @@
Computation times
=================
-**00:34.361** total execution time for **how_to_optimize_operators** files:
+**00:35.649** total execution time for **how_to_optimize_operators** files:
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:32.054 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:33.063 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.244 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.410 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.062 | 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 7c1e5f356..217daa240 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
=================
-**06:23.136** total execution time for **how_to_tune_with_autoscheduler** files:
+**06:35.777** 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``) | 03:17.760 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 03:24.880 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:23.325 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:24.368 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 00:47.730 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 00:56.966 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:36.545 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:31.401 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:08.980 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:09.149 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:08.797 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:09.012 | 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 57772b1dd..496395894 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
@@ -771,7 +771,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 0.367 ms
+ Execution time of this operator: 0.353 ms
@@ -1378,7 +1378,7 @@ In the example below we resume the status and do more 5 trials.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 3 minutes 17.760 seconds)
+ **Total running time of the script:** ( 3 minutes 24.880 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_arm.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_arm.rst.txt
index 58fb7cb5f..338b0b3af 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_arm.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_arm.rst.txt
@@ -333,196 +333,196 @@ The task scheduler will just optimize this objective.
Extract tasks...
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- ========== Task 0 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 1024], [1, 1, 1024, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 1024]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 1024, 1024]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 1024]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 0 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 1024], [1, 1, 1024, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 1024]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 1024, 1024]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 1 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 256], [1, 1, 256, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 1 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 256], [1, 1, 256, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 2 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 256]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 256, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 2 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 256]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 256, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 3 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 128], [1, 1, 128, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 128, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 3 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 128], [1, 1, 128, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 128, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 4 (workload key: ["d7b65649a4dd54becea0a52aabbc5af5", [1, 1000], [1, 1000]]) ==========
- placeholder = PLACEHOLDER [1, 1000]
- T_softmax_maxelem(i0) max= placeholder[i0, k]
- T_softmax_exp(i0, i1) = tir.exp((placeholder[i0, i1] - T_softmax_maxelem[i0]))
+ ========== Task 4 (workload key: ["7d79c516e212fe1d73f5dbb90eaca2cf", [1, 1000], [1, 1000]]) ==========
+ p0 = PLACEHOLDER [1, 1000]
+ T_softmax_maxelem(i0) max= p0[i0, k]
+ T_softmax_exp(i0, i1) = tir.exp((p0[i0, i1] - T_softmax_maxelem[i0]))
T_softmax_expsum(i0) += T_softmax_exp[i0, k]
T_softmax_norm(i0, i1) = (T_softmax_exp[i0, i1]/T_softmax_expsum[i0])
- ========== Task 5 (workload key: ["69115f188984ae34ede37c3b8ca40b43", [1, 7, 7, 1024], [1, 1, 1, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 1024]
- tensor(ax0, ax1, ax2, ax3) += placeholder[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
+ ========== Task 5 (workload key: ["be3babb9a46e32f66b717a3e2a2d522c", [1, 7, 7, 1024], [1, 1, 1, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 1024]
+ tensor(ax0, ax1, ax2, ax3) += p0[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
tensor(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3]/(float32((select((bool)1, ((ax1 + 1)*7), (((ax1 + 1)*7) + 1)) - (ax1*7)))*float32((select((bool)1, ((ax2 + 1)*7), (((ax2 + 1)*7) + 1)) - (ax2*7)))))
- ========== Task 6 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 512], [1, 1, 512, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 1024]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 1024]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 6 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 512], [1, 1, 512, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 1024]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 7 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 128], [1, 1, 128, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 128]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 128, 128]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 7 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 128], [1, 1, 128, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 128]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 128, 128]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 8 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 128]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 128, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 8 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 128]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 128, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 9 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 256]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 256, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 9 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 256]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 256, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 10 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 512]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 512, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 10 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 512]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 512, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 11 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 112, 112, 64], [3, 3, 64, 1], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 112, 112, 64]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 64, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 11 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 112, 112, 64], [3, 3, 64, 1], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 112, 112, 64]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 64, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 12 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 64], [1, 1, 64, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 64, 128]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 12 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 64], [1, 1, 64, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 64, 128]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 13 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 128]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 128, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 13 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 128]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 128, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 14 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 512], [1, 1, 512, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 14 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 512], [1, 1, 512, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 15 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 112, 112, 32], [3, 3, 32, 1], [1, 1, 1, 32], [1, 112, 112, 32]]) ==========
- placeholder = PLACEHOLDER [1, 112, 112, 32]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 32, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 32]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 15 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 112, 112, 32], [3, 3, 32, 1], [1, 1, 1, 32], [1, 112, 112, 32]]) ==========
+ p0 = PLACEHOLDER [1, 112, 112, 32]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 32, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 32]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 16 (workload key: ["2ca148ecea6508ce625f85719021344f", [1, 224, 224, 3], [3, 3, 3, 32], [1, 112, 1, 1], [1, 112, 1, 1], [1, 112, 112, 32]]) ==========
- placeholder = PLACEHOLDER [1, 224, 224, 3]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 225)) && (i2 >= 1)) && (i2 < 225)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 3, 32]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 112, 1, 1]
- T_multiply(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3]*placeholder[ax0, ax1, 0, 0])
- placeholder = PLACEHOLDER [1, 112, 1, 1]
- T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, 0, 0])
+ ========== Task 16 (workload key: ["ad24d4d2f83975ff580a4833fbf3ef94", [1, 224, 224, 3], [3, 3, 3, 32], [1, 112, 1, 1], [1, 112, 1, 1], [1, 112, 112, 32]]) ==========
+ p0 = PLACEHOLDER [1, 224, 224, 3]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 225)) && (i2 >= 1)) && (i2 < 225)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 3, 32]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 112, 1, 1]
+ T_multiply(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3]*p2[ax0, ax1, 0, 0])
+ p3 = PLACEHOLDER [1, 112, 1, 1]
+ T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + p3[ax0, ax1, 0, 0])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 17 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 256], [1, 1, 256, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 17 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 256], [1, 1, 256, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 18 (workload key: ["7d44c6e3c81cd80f61ff2265b2bae89a", [1, 1024], [1000, 1024], [1, 1000], [1, 1000]]) ==========
- placeholder = PLACEHOLDER [1, 1024]
- placeholder = PLACEHOLDER [1000, 1024]
- T_matmul_NT(i, j) += (placeholder[i, k]*placeholder[j, k])
- placeholder = PLACEHOLDER [1, 1000]
- T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + placeholder[ax0, ax1])
-
- ========== Task 19 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 512]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 512, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 18 (workload key: ["00a059b856ac30ac172b6252254479a6", [1, 1024], [1000, 1024], [1, 1000], [1, 1000]]) ==========
+ p0 = PLACEHOLDER [1, 1024]
+ p1 = PLACEHOLDER [1000, 1024]
+ T_matmul_NT(i, j) += (p0[i, k]*p1[j, k])
+ p2 = PLACEHOLDER [1, 1000]
+ T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + p2[ax0, ax1])
+
+ ========== Task 19 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 512]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 512, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 20 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 7, 7, 1024], [3, 3, 1024, 1], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 1024]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 1024, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 1024]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 20 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 7, 7, 1024], [3, 3, 1024, 1], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 1024]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 1024, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 21 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 112, 112, 32], [1, 1, 32, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
- placeholder = PLACEHOLDER [1, 112, 112, 32]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 32, 64]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 21 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 112, 112, 32], [1, 1, 32, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
+ p0 = PLACEHOLDER [1, 112, 112, 32]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 32, 64]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
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 539a0738e..073f7f44d 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
@@ -203,277 +203,277 @@ The task scheduler will just optimize this objective.
Extract tasks...
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- ========== Task 0 (workload key: ["8654f16aeddf785bad9f028164b3a48d", [1, 56, 56, 64], [1, 1, 64, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 64, 64]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-
- ========== Task 1 (workload key: ["c4500b4e2fd04e695c32d2f31bbdc14a", [1, 28, 28, 128], [4, 4, 128, 128], [1, 28, 28, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 0 (workload key: ["2d10de6646307f0e3e5cf4b31c20e69b", [1, 56, 56, 64], [1, 1, 64, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 64, 64]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+
+ ========== Task 1 (workload key: ["f19692ed81d032b1697c08adee62f9a5", [1, 28, 28, 128], [4, 4, 128, 128], [1, 28, 28, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*2) + eps), ((floormod(p, 14)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 128, 128]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 128, 128]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*14)*14) + (floordiv(h, 2)*14)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 28, 28, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+ p3 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 2 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 56, 56, 64], [1, 1, 64, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 64, 128]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
+ ========== Task 2 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 56, 56, 64], [1, 1, 64, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 64, 128]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
- ========== Task 3 (workload key: ["b8b52b9be9df6102466a22a014c44c1f", [1, 14, 14, 256], [4, 4, 256, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 3 (workload key: ["1097323f3970e5c881ad3a0028ca79cb", [1, 14, 14, 256], [4, 4, 256, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 49), ((floormod(floordiv(p, 7), 7)*2) + eps), ((floormod(p, 7)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 256, 256]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 256, 256]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*7)*7) + (floordiv(h, 2)*7)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 4 (workload key: ["e4cdf917b876dbdd64488c3818d9c141", [1, 28, 28, 128], [4, 4, 128, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 4 (workload key: ["0bcf718c0e6566bcd6c3b1437a3b6291", [1, 28, 28, 128], [4, 4, 128, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*2) + eps), ((floormod(p, 14)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 128, 128]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 128, 128]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*14)*14) + (floordiv(h, 2)*14)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 5 (workload key: ["d730bcd28f0920f6b97245e2a11bd8d6", [1, 7, 7, 512], [4, 4, 512, 512], [1, 7, 7, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 5 (workload key: ["d78e8eb6021c4cdda0ad7775d10f751a", [1, 7, 7, 512], [4, 4, 512, 512], [1, 7, 7, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 16), ((floormod(floordiv(p, 4), 4)*2) + eps), ((floormod(p, 4)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 512, 512]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 512, 512]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*4)*4) + (floordiv(h, 2)*4)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
+ p2 = PLACEHOLDER [1, 7, 7, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
- ========== Task 6 (workload key: ["b818b53148cd450f86569dfc3e04cb8a", [1, 56, 56, 64], [6, 6, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 6 (workload key: ["7c2a4f1f432f81c44985590780dfb52d", [1, 56, 56, 64], [6, 6, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*4) + eps), ((floormod(p, 14)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [6, 6, 64, 64]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [6, 6, 64, 64]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*14)*14) + (floordiv(h, 4)*14)) + floordiv(w, 4)), co]
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 7 (workload key: ["ad6cecbf5d85cb1cda3c2bb7af170211", [1, 7, 7, 512], [4, 4, 512, 512], [1, 7, 7, 512], [1, 1, 1, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 7 (workload key: ["a3df19e5b88592ef5a9ce584a1ca3010", [1, 7, 7, 512], [4, 4, 512, 512], [1, 7, 7, 512], [1, 1, 1, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 16), ((floormod(floordiv(p, 4), 4)*2) + eps), ((floormod(p, 4)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 512, 512]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 512, 512]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*4)*4) + (floordiv(h, 2)*4)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_multiply(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3]*placeholder[ax0, 0, 0, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 7, 7, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+ p3 = PLACEHOLDER [1, 1, 1, 512]
+ T_multiply(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3]*p3[ax0, 0, 0, ax3])
+ p4 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + p4[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 8 (workload key: ["f3b6c10fcc6ce01ff01add933e4d21e9", [1, 14, 14, 256], [4, 4, 256, 256], [1, 14, 14, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 8 (workload key: ["64b7ce5264a64cb340d78b444b0325e6", [1, 14, 14, 256], [4, 4, 256, 256], [1, 14, 14, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 49), ((floormod(floordiv(p, 7), 7)*2) + eps), ((floormod(p, 7)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 256, 256]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 256, 256]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*7)*7) + (floordiv(h, 2)*7)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 14, 14, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+ p3 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 9 (workload key: ["d7b65649a4dd54becea0a52aabbc5af5", [1, 1000], [1, 1000]]) ==========
- placeholder = PLACEHOLDER [1, 1000]
- T_softmax_maxelem(i0) max= placeholder[i0, k]
- T_softmax_exp(i0, i1) = tir.exp((placeholder[i0, i1] - T_softmax_maxelem[i0]))
+ ========== Task 9 (workload key: ["7d79c516e212fe1d73f5dbb90eaca2cf", [1, 1000], [1, 1000]]) ==========
+ p0 = PLACEHOLDER [1, 1000]
+ T_softmax_maxelem(i0) max= p0[i0, k]
+ T_softmax_exp(i0, i1) = tir.exp((p0[i0, i1] - T_softmax_maxelem[i0]))
T_softmax_expsum(i0) += T_softmax_exp[i0, k]
T_softmax_norm(i0, i1) = (T_softmax_exp[i0, i1]/T_softmax_expsum[i0])
- ========== Task 10 (workload key: ["69115f188984ae34ede37c3b8ca40b43", [1, 7, 7, 512], [1, 1, 1, 512]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- tensor(ax0, ax1, ax2, ax3) += placeholder[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
+ ========== Task 10 (workload key: ["be3babb9a46e32f66b717a3e2a2d522c", [1, 7, 7, 512], [1, 1, 1, 512]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ tensor(ax0, ax1, ax2, ax3) += p0[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
tensor(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3]/(float32((select((bool)1, ((ax1 + 1)*7), (((ax1 + 1)*7) + 1)) - (ax1*7)))*float32((select((bool)1, ((ax2 + 1)*7), (((ax2 + 1)*7) + 1)) - (ax2*7)))))
- ========== Task 11 (workload key: ["3a69f9fbc63760d99e36b4c17b3bfc57", [1, 7, 7, 512], [4, 4, 512, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 11 (workload key: ["40b1cf1fd37b0ef111b3cc0247302508", [1, 7, 7, 512], [4, 4, 512, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 16), ((floormod(floordiv(p, 4), 4)*2) + eps), ((floormod(p, 4)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 512, 512]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 512, 512]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*4)*4) + (floordiv(h, 2)*4)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 12 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 28, 28, 128], [1, 1, 128, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 128, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-
- ========== Task 13 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 14, 14, 256], [3, 3, 256, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 256, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 12 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 28, 28, 128], [1, 1, 128, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 128, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+
+ ========== Task 13 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 14, 14, 256], [3, 3, 256, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 256, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 14 (workload key: ["dac19035dd5fe9424ee8617421b9c817", [1, 28, 28, 128], [4, 4, 128, 128], [1, 28, 28, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 14 (workload key: ["25577781e50c611c2e45e73c1cb3a6ca", [1, 28, 28, 128], [4, 4, 128, 128], [1, 28, 28, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*2) + eps), ((floormod(p, 14)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 128, 128]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 128, 128]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*14)*14) + (floordiv(h, 2)*14)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
- ========== Task 15 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 28, 28, 128], [3, 3, 128, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 128, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 28, 28, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+ ========== Task 15 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 28, 28, 128], [3, 3, 128, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 128, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 16 (workload key: ["1e3c4211ffd2f2db91078ae4d04b779d", [1, 56, 56, 64], [6, 6, 64, 64], [1, 56, 56, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 16 (workload key: ["6c4f6234946e16bcf9e48bdf289f9200", [1, 56, 56, 64], [6, 6, 64, 64], [1, 56, 56, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*4) + eps), ((floormod(p, 14)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [6, 6, 64, 64]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [6, 6, 64, 64]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*14)*14) + (floordiv(h, 4)*14)) + floordiv(w, 4)), co]
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 56, 56, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+ p3 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 17 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 224, 224, 3], [7, 7, 3, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
- placeholder = PLACEHOLDER [1, 224, 224, 3]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 3) && (i1 < 227)) && (i2 >= 3)) && (i2 < 227)), placeholder[i0, (i1 - 3), (i2 - 3), i3], 0f)
- placeholder = PLACEHOLDER [7, 7, 3, 64]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 17 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 224, 224, 3], [7, 7, 3, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
+ p0 = PLACEHOLDER [1, 224, 224, 3]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 3) && (i1 < 227)) && (i2 >= 3)) && (i2 < 227)), p0[i0, (i1 - 3), (i2 - 3), i3], 0f)
+ p1 = PLACEHOLDER [7, 7, 3, 64]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 18 (workload key: ["3ea73fb9b0364374730d09e068821f95", [1, 56, 56, 64], [6, 6, 64, 64], [1, 56, 56, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 18 (workload key: ["10b8215aaf2e14d47d40b4093e6f41a0", [1, 56, 56, 64], [6, 6, 64, 64], [1, 56, 56, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*4) + eps), ((floormod(p, 14)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [6, 6, 64, 64]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [6, 6, 64, 64]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*14)*14) + (floordiv(h, 4)*14)) + floordiv(w, 4)), co]
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
+ p2 = PLACEHOLDER [1, 56, 56, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
- ========== Task 19 (workload key: ["d374e472bd9d8164892b9e28a0a8cb59", [1, 14, 14, 256], [4, 4, 256, 256], [1, 14, 14, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 19 (workload key: ["7f3fee61bc3c2604395f5d343b840b7c", [1, 14, 14, 256], [4, 4, 256, 256], [1, 14, 14, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 49), ((floormod(floordiv(p, 7), 7)*2) + eps), ((floormod(p, 7)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [4, 4, 256, 256]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [4, 4, 256, 256]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*7)*7) + (floordiv(h, 2)*7)) + floordiv(w, 2)), co]
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
+ p2 = PLACEHOLDER [1, 14, 14, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
- ========== Task 20 (workload key: ["64b98c71af70a904fdbb81d7d4188d84", [1, 112, 112, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 112, 112, 64]
- pad_temp(ax0, ax1, ax2, ax3) = tir.if_then_else(((((ax1 >= 1) && (ax1 < 113)) && (ax2 >= 1)) && (ax2 < 113)), placeholder[ax0, (ax1 - 1), (ax2 - 1), ax3], -3.40282e+38f)
+ ========== Task 20 (workload key: ["affd3c4a65f665e451a06d65bf32750d", [1, 112, 112, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 112, 112, 64]
+ pad_temp(ax0, ax1, ax2, ax3) = tir.if_then_else(((((ax1 >= 1) && (ax1 < 113)) && (ax2 >= 1)) && (ax2 < 113)), p0[ax0, (ax1 - 1), (ax2 - 1), ax3], -3.40282e+38f)
tensor(ax0, ax1, ax2, ax3) max= pad_temp[ax0, ((ax1*2) + rv0), ((ax2*2) + rv1), ax3]
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p1 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3] + p1[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 21 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 14, 14, 256], [1, 1, 256, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-
- ========== Task 22 (workload key: ["7d44c6e3c81cd80f61ff2265b2bae89a", [1, 512], [1000, 512], [1, 1000], [1, 1000]]) ==========
- placeholder = PLACEHOLDER [1, 512]
- placeholder = PLACEHOLDER [1000, 512]
- T_matmul_NT(i, j) += (placeholder[i, k]*placeholder[j, k])
- placeholder = PLACEHOLDER [1, 1000]
- T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + placeholder[ax0, ax1])
-
- ========== Task 23 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 56, 56, 64], [3, 3, 64, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 64, 128]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 21 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 14, 14, 256], [1, 1, 256, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+
+ ========== Task 22 (workload key: ["00a059b856ac30ac172b6252254479a6", [1, 512], [1000, 512], [1, 1000], [1, 1000]]) ==========
+ p0 = PLACEHOLDER [1, 512]
+ p1 = PLACEHOLDER [1000, 512]
+ T_matmul_NT(i, j) += (p0[i, k]*p1[j, k])
+ p2 = PLACEHOLDER [1, 1000]
+ T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + p2[ax0, ax1])
+
+ ========== Task 23 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 56, 56, 64], [3, 3, 64, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 64, 128]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
@@ -647,7 +647,7 @@ so we can read the log file and load the best schedules.
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 10.0240 10.0604 10.0697 9.9419 0.0582
+ 8.1562 8.1569 8.1608 8.1510 0.0040
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_mali.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_mali.rst.txt
index 595793a9a..406de3e35 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_mali.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_mali.rst.txt
@@ -232,196 +232,196 @@ The task scheduler will just optimize this objective.
Extract tasks...
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- ========== Task 0 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 1024], [1, 1, 1024, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 1024]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 1024, 1024]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 1024]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 0 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 1024], [1, 1, 1024, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 1024]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 1024, 1024]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 1 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 256], [1, 1, 256, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 1 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 256], [1, 1, 256, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 2 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 256]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 256, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 2 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 256]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 256, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 3 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 128], [1, 1, 128, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 128, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 3 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 128], [1, 1, 128, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 128, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 4 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 512], [1, 1, 512, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 1024]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 1024]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 4 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 512], [1, 1, 512, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 1024]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 5 (workload key: ["69115f188984ae34ede37c3b8ca40b43", [1, 7, 7, 1024], [1, 1, 1, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 1024]
- tensor(ax0, ax1, ax2, ax3) += placeholder[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
+ ========== Task 5 (workload key: ["be3babb9a46e32f66b717a3e2a2d522c", [1, 7, 7, 1024], [1, 1, 1, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 1024]
+ tensor(ax0, ax1, ax2, ax3) += p0[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
tensor(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3]/(float32((select((bool)1, ((ax1 + 1)*7), (((ax1 + 1)*7) + 1)) - (ax1*7)))*float32((select((bool)1, ((ax2 + 1)*7), (((ax2 + 1)*7) + 1)) - (ax2*7)))))
- ========== Task 6 (workload key: ["d7b65649a4dd54becea0a52aabbc5af5", [1, 1000], [1, 1000]]) ==========
- placeholder = PLACEHOLDER [1, 1000]
- T_softmax_maxelem(i0) max= placeholder[i0, k]
- T_softmax_exp(i0, i1) = tir.exp((placeholder[i0, i1] - T_softmax_maxelem[i0]))
+ ========== Task 6 (workload key: ["7d79c516e212fe1d73f5dbb90eaca2cf", [1, 1000], [1, 1000]]) ==========
+ p0 = PLACEHOLDER [1, 1000]
+ T_softmax_maxelem(i0) max= p0[i0, k]
+ T_softmax_exp(i0, i1) = tir.exp((p0[i0, i1] - T_softmax_maxelem[i0]))
T_softmax_expsum(i0) += T_softmax_exp[i0, k]
T_softmax_norm(i0, i1) = (T_softmax_exp[i0, i1]/T_softmax_expsum[i0])
- ========== Task 7 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 128], [1, 1, 128, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 128]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 128, 128]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 7 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 128], [1, 1, 128, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 128]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 128, 128]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 8 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 7, 7, 1024], [3, 3, 1024, 1], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 1024]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 1024, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 1024]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 8 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 7, 7, 1024], [3, 3, 1024, 1], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 1024]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 1024, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 9 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 64], [1, 1, 64, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 64, 128]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 9 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 64], [1, 1, 64, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 64, 128]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 10 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 512]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 512, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 10 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 512]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 512, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 11 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 256], [1, 1, 256, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 11 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 256], [1, 1, 256, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 12 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 128]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 128, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 12 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 128]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 128, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 13 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 112, 112, 64], [3, 3, 64, 1], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 112, 112, 64]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 64, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 13 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 112, 112, 64], [3, 3, 64, 1], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 112, 112, 64]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 64, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 14 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 112, 112, 32], [3, 3, 32, 1], [1, 1, 1, 32], [1, 112, 112, 32]]) ==========
- placeholder = PLACEHOLDER [1, 112, 112, 32]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 32, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 32]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 14 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 112, 112, 32], [3, 3, 32, 1], [1, 1, 1, 32], [1, 112, 112, 32]]) ==========
+ p0 = PLACEHOLDER [1, 112, 112, 32]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 32, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 32]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 15 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 128]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 128, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 15 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 128]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 128, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 16 (workload key: ["2ca148ecea6508ce625f85719021344f", [1, 224, 224, 3], [3, 3, 3, 32], [1, 112, 1, 1], [1, 112, 1, 1], [1, 112, 112, 32]]) ==========
- placeholder = PLACEHOLDER [1, 224, 224, 3]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 225)) && (i2 >= 1)) && (i2 < 225)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 3, 32]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 112, 1, 1]
- T_multiply(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3]*placeholder[ax0, ax1, 0, 0])
- placeholder = PLACEHOLDER [1, 112, 1, 1]
- T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, 0, 0])
+ ========== Task 16 (workload key: ["ad24d4d2f83975ff580a4833fbf3ef94", [1, 224, 224, 3], [3, 3, 3, 32], [1, 112, 1, 1], [1, 112, 1, 1], [1, 112, 112, 32]]) ==========
+ p0 = PLACEHOLDER [1, 224, 224, 3]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 225)) && (i2 >= 1)) && (i2 < 225)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 3, 32]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 112, 1, 1]
+ T_multiply(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3]*p2[ax0, ax1, 0, 0])
+ p3 = PLACEHOLDER [1, 112, 1, 1]
+ T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + p3[ax0, ax1, 0, 0])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 17 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 512], [1, 1, 512, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 17 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 512], [1, 1, 512, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 18 (workload key: ["7d44c6e3c81cd80f61ff2265b2bae89a", [1, 1024], [1000, 1024], [1, 1000], [1, 1000]]) ==========
- placeholder = PLACEHOLDER [1, 1024]
- placeholder = PLACEHOLDER [1000, 1024]
- T_matmul_NT(i, j) += (placeholder[i, k]*placeholder[j, k])
- placeholder = PLACEHOLDER [1, 1000]
- T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + placeholder[ax0, ax1])
-
- ========== Task 19 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 512]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 512, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 18 (workload key: ["00a059b856ac30ac172b6252254479a6", [1, 1024], [1000, 1024], [1, 1000], [1, 1000]]) ==========
+ p0 = PLACEHOLDER [1, 1024]
+ p1 = PLACEHOLDER [1000, 1024]
+ T_matmul_NT(i, j) += (p0[i, k]*p1[j, k])
+ p2 = PLACEHOLDER [1, 1000]
+ T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + p2[ax0, ax1])
+
+ ========== Task 19 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 512]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 512, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 20 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 256]
- PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 256, 1]
- DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 20 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 256]
+ PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 256, 1]
+ DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 21 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 112, 112, 32], [1, 1, 32, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
- placeholder = PLACEHOLDER [1, 112, 112, 32]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 32, 64]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 21 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 112, 112, 32], [1, 1, 32, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
+ p0 = PLACEHOLDER [1, 112, 112, 32]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 32, 64]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
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 c259ed741..6cd80747a 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
@@ -226,262 +226,262 @@ The task scheduler will just optimize this objective.
Extract tasks...
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- ========== Task 0 (workload key: ["8654f16aeddf785bad9f028164b3a48d", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 64, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-
- ========== Task 1 (workload key: ["12cb81d4ad0a81be02dedf09d1ac8391", [1, 14, 14, 256], [1, 1, 256, 1024], [1, 14, 14, 1024], [1, 14, 14, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 1024]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 14, 14, 1024]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
- ========== Task 2 (workload key: ["b9a4f9bd1416ba25810cb3de27628ace", [1, 14, 14, 256], [1, 1, 256, 1024], [1, 14, 14, 1024], [1, 1, 1, 1024], [1, 14, 14, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 1024]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 14, 14, 1024]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 1024]
- T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 0 (workload key: ["2d10de6646307f0e3e5cf4b31c20e69b", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 64, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+
+ ========== Task 1 (workload key: ["3060808fc5c74e18b1276729071fbae0", [1, 14, 14, 256], [1, 1, 256, 1024], [1, 14, 14, 1024], [1, 14, 14, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 1024]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 14, 14, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+ ========== Task 2 (workload key: ["76afb7bf408a1ffa0b8b7bc09d077dc3", [1, 14, 14, 256], [1, 1, 256, 1024], [1, 14, 14, 1024], [1, 1, 1, 1024], [1, 14, 14, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 1024]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 14, 14, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+ p3 = PLACEHOLDER [1, 1, 1, 1024]
+ T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 3 (workload key: ["56af7508fbcdf6d851892b1e8434667b", [1, 14, 14, 1024], [1, 1, 1024, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 1024]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 1024, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 3 (workload key: ["2beb39e9afe4c74822fffbcbb8533595", [1, 14, 14, 1024], [1, 1, 1024, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 1024]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 1024, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 4 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 56, 56, 256], [1, 1, 256, 512], [1, 28, 28, 512]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
+ ========== Task 4 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 56, 56, 256], [1, 1, 256, 512], [1, 28, 28, 512]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
- ========== Task 5 (workload key: ["c68f92478eb18145106184c587d212b6", [1, 14, 14, 256], [6, 6, 256, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 256]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 5 (workload key: ["38552500208b25b4035682b0e93cbce3", [1, 14, 14, 256], [6, 6, 256, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 256]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 16), ((floormod(floordiv(p, 4), 4)*4) + eps), ((floormod(p, 4)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [6, 6, 256, 256]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [6, 6, 256, 256]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*4)*4) + (floordiv(h, 4)*4)) + floordiv(w, 4)), co]
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 6 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 1024], [1, 1, 1024, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 1024]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 1024, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 6 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 1024], [1, 1, 1024, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 1024]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 1024, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 7 (workload key: ["12cb81d4ad0a81be02dedf09d1ac8391", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256], [1, 56, 56, 256]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 64, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 56, 56, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
- ========== Task 8 (workload key: ["ecec634b4882c5731f86cce3109db636", [1, 28, 28, 128], [6, 6, 128, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ ========== Task 7 (workload key: ["3060808fc5c74e18b1276729071fbae0", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256], [1, 56, 56, 256]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 64, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 56, 56, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+ ========== Task 8 (workload key: ["cfd09cf1ca9e943f0ee12a18813a5c75", [1, 28, 28, 128], [6, 6, 128, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 49), ((floormod(floordiv(p, 7), 7)*4) + eps), ((floormod(p, 7)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
- placeholder = PLACEHOLDER [6, 6, 128, 128]
- bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+ p1 = PLACEHOLDER [6, 6, 128, 128]
+ bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*7)*7) + (floordiv(h, 4)*7)) + floordiv(w, 4)), co]
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 9 (workload key: ["12cb81d4ad0a81be02dedf09d1ac8391", [1, 28, 28, 128], [1, 1, 128, 512], [1, 28, 28, 512], [1, 28, 28, 512]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 128, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 28, 28, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
- ========== Task 10 (workload key: ["d7b65649a4dd54becea0a52aabbc5af5", [1, 1000], [1, 1000]]) ==========
- placeholder = PLACEHOLDER [1, 1000]
- T_softmax_maxelem(i0) max= placeholder[i0, k]
- T_softmax_exp(i0, i1) = tir.exp((placeholder[i0, i1] - T_softmax_maxelem[i0]))
+ ========== Task 9 (workload key: ["3060808fc5c74e18b1276729071fbae0", [1, 28, 28, 128], [1, 1, 128, 512], [1, 28, 28, 512], [1, 28, 28, 512]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 128, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 28, 28, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+ ========== Task 10 (workload key: ["7d79c516e212fe1d73f5dbb90eaca2cf", [1, 1000], [1, 1000]]) ==========
+ p0 = PLACEHOLDER [1, 1000]
+ T_softmax_maxelem(i0) max= p0[i0, k]
+ T_softmax_exp(i0, i1) = tir.exp((p0[i0, i1] - T_softmax_maxelem[i0]))
T_softmax_expsum(i0) += T_softmax_exp[i0, k]
T_softmax_norm(i0, i1) = (T_softmax_exp[i0, i1]/T_softmax_expsum[i0])
- ========== Task 11 (workload key: ["69115f188984ae34ede37c3b8ca40b43", [1, 7, 7, 2048], [1, 1, 1, 2048]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 2048]
- tensor(ax0, ax1, ax2, ax3) += placeholder[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
+ ========== Task 11 (workload key: ["be3babb9a46e32f66b717a3e2a2d522c", [1, 7, 7, 2048], [1, 1, 1, 2048]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 2048]
+ tensor(ax0, ax1, ax2, ax3) += p0[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
tensor(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3]/(float32((select((bool)1, ((ax1 + 1)*7), (((ax1 + 1)*7) + 1)) - (ax1*7)))*float32((select((bool)1, ((ax2 + 1)*7), (((ax2 + 1)*7) + 1)) - (ax2*7)))))
- ========== Task 12 (workload key: ["12cb81d4ad0a81be02dedf09d1ac8391", [1, 7, 7, 512], [1, 1, 512, 2048], [1, 7, 7, 2048], [1, 7, 7, 2048]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 2048]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 7, 7, 2048]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
- ========== Task 13 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 512], [1, 1, 512, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 128]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 12 (workload key: ["3060808fc5c74e18b1276729071fbae0", [1, 7, 7, 512], [1, 1, 512, 2048], [1, 7, 7, 2048], [1, 7, 7, 2048]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 2048]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 7, 7, 2048]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+ ========== Task 13 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 512], [1, 1, 512, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 128]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 14 (workload key: ["56af7508fbcdf6d851892b1e8434667b", [1, 28, 28, 512], [1, 1, 512, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 14 (workload key: ["2beb39e9afe4c74822fffbcbb8533595", [1, 28, 28, 512], [1, 1, 512, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 15 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 28, 28, 512], [1, 1, 512, 1024], [1, 14, 14, 1024]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 1024]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-
- ========== Task 16 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 2048], [1, 1, 2048, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 2048]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 2048, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 15 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 28, 28, 512], [1, 1, 512, 1024], [1, 14, 14, 1024]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 1024]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+
+ ========== Task 16 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 2048], [1, 1, 2048, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 2048]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 2048, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 17 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 256], [1, 1, 256, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 64]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 17 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 256], [1, 1, 256, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 64]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 18 (workload key: ["b9a4f9bd1416ba25810cb3de27628ace", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256], [1, 1, 1, 256], [1, 56, 56, 256]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 64, 256]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 56, 56, 256]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 256]
- T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 18 (workload key: ["76afb7bf408a1ffa0b8b7bc09d077dc3", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256], [1, 1, 1, 256], [1, 56, 56, 256]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 64, 256]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 56, 56, 256]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+ p3 = PLACEHOLDER [1, 1, 1, 256]
+ T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 19 (workload key: ["b9a4f9bd1416ba25810cb3de27628ace", [1, 28, 28, 128], [1, 1, 128, 512], [1, 28, 28, 512], [1, 1, 1, 512], [1, 28, 28, 512]]) ==========
- placeholder = PLACEHOLDER [1, 28, 28, 128]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 128, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 28, 28, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 19 (workload key: ["76afb7bf408a1ffa0b8b7bc09d077dc3", [1, 28, 28, 128], [1, 1, 128, 512], [1, 28, 28, 512], [1, 1, 1, 512], [1, 28, 28, 512]]) ==========
+ p0 = PLACEHOLDER [1, 28, 28, 128]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 128, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 28, 28, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+ p3 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 20 (workload key: ["86551f1a74663d3ceafd5884659d3478", [1, 7, 7, 512], [3, 3, 512, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 512, 512]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 512]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 20 (workload key: ["d37380659057397544e056461ea3bad3", [1, 7, 7, 512], [3, 3, 512, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 512, 512]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 512]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 21 (workload key: ["86551f1a74663d3ceafd5884659d3478", [1, 56, 56, 64], [3, 3, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
- placeholder = PLACEHOLDER [3, 3, 64, 64]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 21 (workload key: ["d37380659057397544e056461ea3bad3", [1, 56, 56, 64], [3, 3, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+ p1 = PLACEHOLDER [3, 3, 64, 64]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 22 (workload key: ["56af7508fbcdf6d851892b1e8434667b", [1, 56, 56, 256], [1, 1, 256, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 256]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 256, 128]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 128]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 22 (workload key: ["2beb39e9afe4c74822fffbcbb8533595", [1, 56, 56, 256], [1, 1, 256, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 256]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 256, 128]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 128]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 23 (workload key: ["ff5ea7f814e5c497bb685e7385cf7159", [1, 7, 7, 512], [1, 1, 512, 2048], [1, 7, 7, 2048], [1, 1, 1, 2048], [1, 1, 1, 2048], [1, 7, 7, 2048]]) ==========
- placeholder = PLACEHOLDER [1, 7, 7, 512]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 512, 2048]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 7, 7, 2048]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 2048]
- T_multiply(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3]*placeholder[ax0, 0, 0, ax3])
- placeholder = PLACEHOLDER [1, 1, 1, 2048]
- T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 23 (workload key: ["f07e228ef5f642b386d23a62df615e7b", [1, 7, 7, 512], [1, 1, 512, 2048], [1, 7, 7, 2048], [1, 1, 1, 2048], [1, 1, 1, 2048], [1, 7, 7, 2048]]) ==========
+ p0 = PLACEHOLDER [1, 7, 7, 512]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 512, 2048]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 7, 7, 2048]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+ p3 = PLACEHOLDER [1, 1, 1, 2048]
+ T_multiply(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3]*p3[ax0, 0, 0, ax3])
+ p4 = PLACEHOLDER [1, 1, 1, 2048]
+ T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + p4[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 24 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 224, 224, 3], [7, 7, 3, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
- placeholder = PLACEHOLDER [1, 224, 224, 3]
- pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 3) && (i1 < 227)) && (i2 >= 3)) && (i2 < 227)), placeholder[i0, (i1 - 3), (i2 - 3), i3], 0f)
- placeholder = PLACEHOLDER [7, 7, 3, 64]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 24 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 224, 224, 3], [7, 7, 3, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
+ p0 = PLACEHOLDER [1, 224, 224, 3]
+ pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 3) && (i1 < 227)) && (i2 >= 3)) && (i2 < 227)), p0[i0, (i1 - 3), (i2 - 3), i3], 0f)
+ p1 = PLACEHOLDER [7, 7, 3, 64]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 25 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 14, 14, 1024], [1, 1, 1024, 2048], [1, 7, 7, 2048]]) ==========
- placeholder = PLACEHOLDER [1, 14, 14, 1024]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 1024, 2048]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-
- ========== Task 26 (workload key: ["7d44c6e3c81cd80f61ff2265b2bae89a", [1, 2048], [1000, 2048], [1, 1000], [1, 1000]]) ==========
- placeholder = PLACEHOLDER [1, 2048]
- placeholder = PLACEHOLDER [1000, 2048]
- T_matmul_NT(i, j) += (placeholder[i, k]*placeholder[j, k])
- placeholder = PLACEHOLDER [1, 1000]
- T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + placeholder[ax0, ax1])
-
- ========== Task 27 (workload key: ["64b98c71af70a904fdbb81d7d4188d84", [1, 112, 112, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 112, 112, 64]
- pad_temp(ax0, ax1, ax2, ax3) = tir.if_then_else(((((ax1 >= 1) && (ax1 < 113)) && (ax2 >= 1)) && (ax2 < 113)), placeholder[ax0, (ax1 - 1), (ax2 - 1), ax3], -3.40282e+38f)
+ ========== Task 25 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 14, 14, 1024], [1, 1, 1024, 2048], [1, 7, 7, 2048]]) ==========
+ p0 = PLACEHOLDER [1, 14, 14, 1024]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 1024, 2048]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+
+ ========== Task 26 (workload key: ["00a059b856ac30ac172b6252254479a6", [1, 2048], [1000, 2048], [1, 1000], [1, 1000]]) ==========
+ p0 = PLACEHOLDER [1, 2048]
+ p1 = PLACEHOLDER [1000, 2048]
+ T_matmul_NT(i, j) += (p0[i, k]*p1[j, k])
+ p2 = PLACEHOLDER [1, 1000]
+ T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + p2[ax0, ax1])
+
+ ========== Task 27 (workload key: ["affd3c4a65f665e451a06d65bf32750d", [1, 112, 112, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 112, 112, 64]
+ pad_temp(ax0, ax1, ax2, ax3) = tir.if_then_else(((((ax1 >= 1) && (ax1 < 113)) && (ax2 >= 1)) && (ax2 < 113)), p0[ax0, (ax1 - 1), (ax2 - 1), ax3], -3.40282e+38f)
tensor(ax0, ax1, ax2, ax3) max= pad_temp[ax0, ((ax1*2) + rv0), ((ax2*2) + rv1), ax3]
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ p1 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3] + p1[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
- ========== Task 28 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 64], [1, 1, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
- placeholder = PLACEHOLDER [1, 56, 56, 64]
- pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
- placeholder = PLACEHOLDER [1, 1, 64, 64]
- conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
- placeholder = PLACEHOLDER [1, 1, 1, 64]
- T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+ ========== Task 28 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 64], [1, 1, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+ p0 = PLACEHOLDER [1, 56, 56, 64]
+ pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+ p1 = PLACEHOLDER [1, 1, 64, 64]
+ conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+ p2 = PLACEHOLDER [1, 1, 1, 64]
+ T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
@@ -666,7 +666,7 @@ so we can read the log file and load the best schedules.
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 757.9706 757.1987 759.9952 756.7179 1.4450
+ 760.1457 760.9216 762.4794 757.0362 2.2889
@@ -694,7 +694,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 23.325 seconds)
+ **Total running time of the script:** ( 1 minutes 24.369 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 9d81d87de..a07642aae 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
@@ -397,792 +397,409 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [65536], []),
compute: Buffer(compute_2: Pointer(float32), float32, [65536], [])}
buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute}
- preflattened_buffer_map = {placeholder_6: placeholder_15: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_7: placeholder_16: Buffer(placeholder_12, int32, [4916], []), placeholder_5: placeholder_17: Buffer(placeholder_10, float32, [128, 256], []), placeholder_9: placeholder_18: Buffer(placeholder_14, float32, [128, 512], []), placeholder_8: placeholder_19: Buffer(placeholder_13, int32, [33], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], [])} {
- for (i0.outer: int32, 0, 8) "parallel" {
- allocate(compute_4: Pointer(global float32), float32, [512]), storage_scope = global;
- for (i1.outer: int32, 0, 16) {
- for (nb_j.inner: int32, 0, 2) {
- let cse_var_2: int32 = (nb_j.inner*16)
- let cse_var_1: int32 = ((i1.outer*2) + nb_j.inner)
- {
- compute_5: Buffer(compute_4, float32, [512], [])[cse_var_2] = 0f32
- compute_5[(cse_var_2 + 1)] = 0f32
- compute_5[(cse_var_2 + 2)] = 0f32
- compute_5[(cse_var_2 + 3)] = 0f32
- compute_5[(cse_var_2 + 4)] = 0f32
- compute_5[(cse_var_2 + 5)] = 0f32
- compute_5[(cse_var_2 + 6)] = 0f32
- compute_5[(cse_var_2 + 7)] = 0f32
- compute_5[(cse_var_2 + 8)] = 0f32
- compute_5[(cse_var_2 + 9)] = 0f32
- compute_5[(cse_var_2 + 10)] = 0f32
- compute_5[(cse_var_2 + 11)] = 0f32
- compute_5[(cse_var_2 + 12)] = 0f32
- compute_5[(cse_var_2 + 13)] = 0f32
- compute_5[(cse_var_2 + 14)] = 0f32
- compute_5[(cse_var_2 + 15)] = 0f32
- compute_5[(cse_var_2 + 32)] = 0f32
- compute_5[(cse_var_2 + 33)] = 0f32
- compute_5[(cse_var_2 + 34)] = 0f32
- compute_5[(cse_var_2 + 35)] = 0f32
- compute_5[(cse_var_2 + 36)] = 0f32
- compute_5[(cse_var_2 + 37)] = 0f32
- compute_5[(cse_var_2 + 38)] = 0f32
- compute_5[(cse_var_2 + 39)] = 0f32
- compute_5[(cse_var_2 + 40)] = 0f32
- compute_5[(cse_var_2 + 41)] = 0f32
- compute_5[(cse_var_2 + 42)] = 0f32
- compute_5[(cse_var_2 + 43)] = 0f32
- compute_5[(cse_var_2 + 44)] = 0f32
- compute_5[(cse_var_2 + 45)] = 0f32
- compute_5[(cse_var_2 + 46)] = 0f32
- compute_5[(cse_var_2 + 47)] = 0f32
- compute_5[(cse_var_2 + 64)] = 0f32
- compute_5[(cse_var_2 + 65)] = 0f32
- compute_5[(cse_var_2 + 66)] = 0f32
- compute_5[(cse_var_2 + 67)] = 0f32
- compute_5[(cse_var_2 + 68)] = 0f32
- compute_5[(cse_var_2 + 69)] = 0f32
- compute_5[(cse_var_2 + 70)] = 0f32
- compute_5[(cse_var_2 + 71)] = 0f32
- compute_5[(cse_var_2 + 72)] = 0f32
- compute_5[(cse_var_2 + 73)] = 0f32
- compute_5[(cse_var_2 + 74)] = 0f32
- compute_5[(cse_var_2 + 75)] = 0f32
- compute_5[(cse_var_2 + 76)] = 0f32
- compute_5[(cse_var_2 + 77)] = 0f32
- compute_5[(cse_var_2 + 78)] = 0f32
- compute_5[(cse_var_2 + 79)] = 0f32
- compute_5[(cse_var_2 + 96)] = 0f32
- compute_5[(cse_var_2 + 97)] = 0f32
- compute_5[(cse_var_2 + 98)] = 0f32
- compute_5[(cse_var_2 + 99)] = 0f32
- compute_5[(cse_var_2 + 100)] = 0f32
- compute_5[(cse_var_2 + 101)] = 0f32
- compute_5[(cse_var_2 + 102)] = 0f32
- compute_5[(cse_var_2 + 103)] = 0f32
- compute_5[(cse_var_2 + 104)] = 0f32
- compute_5[(cse_var_2 + 105)] = 0f32
- compute_5[(cse_var_2 + 106)] = 0f32
- compute_5[(cse_var_2 + 107)] = 0f32
- compute_5[(cse_var_2 + 108)] = 0f32
- compute_5[(cse_var_2 + 109)] = 0f32
- compute_5[(cse_var_2 + 110)] = 0f32
- compute_5[(cse_var_2 + 111)] = 0f32
- compute_5[(cse_var_2 + 128)] = 0f32
- compute_5[(cse_var_2 + 129)] = 0f32
- compute_5[(cse_var_2 + 130)] = 0f32
- compute_5[(cse_var_2 + 131)] = 0f32
- compute_5[(cse_var_2 + 132)] = 0f32
- compute_5[(cse_var_2 + 133)] = 0f32
- compute_5[(cse_var_2 + 134)] = 0f32
- compute_5[(cse_var_2 + 135)] = 0f32
- compute_5[(cse_var_2 + 136)] = 0f32
- compute_5[(cse_var_2 + 137)] = 0f32
- compute_5[(cse_var_2 + 138)] = 0f32
- compute_5[(cse_var_2 + 139)] = 0f32
- compute_5[(cse_var_2 + 140)] = 0f32
- compute_5[(cse_var_2 + 141)] = 0f32
- compute_5[(cse_var_2 + 142)] = 0f32
- compute_5[(cse_var_2 + 143)] = 0f32
- compute_5[(cse_var_2 + 160)] = 0f32
- compute_5[(cse_var_2 + 161)] = 0f32
- compute_5[(cse_var_2 + 162)] = 0f32
- compute_5[(cse_var_2 + 163)] = 0f32
- compute_5[(cse_var_2 + 164)] = 0f32
- compute_5[(cse_var_2 + 165)] = 0f32
- compute_5[(cse_var_2 + 166)] = 0f32
- compute_5[(cse_var_2 + 167)] = 0f32
- compute_5[(cse_var_2 + 168)] = 0f32
- compute_5[(cse_var_2 + 169)] = 0f32
- compute_5[(cse_var_2 + 170)] = 0f32
- compute_5[(cse_var_2 + 171)] = 0f32
- compute_5[(cse_var_2 + 172)] = 0f32
- compute_5[(cse_var_2 + 173)] = 0f32
- compute_5[(cse_var_2 + 174)] = 0f32
- compute_5[(cse_var_2 + 175)] = 0f32
- compute_5[(cse_var_2 + 192)] = 0f32
- compute_5[(cse_var_2 + 193)] = 0f32
- compute_5[(cse_var_2 + 194)] = 0f32
- compute_5[(cse_var_2 + 195)] = 0f32
- compute_5[(cse_var_2 + 196)] = 0f32
- compute_5[(cse_var_2 + 197)] = 0f32
- compute_5[(cse_var_2 + 198)] = 0f32
- compute_5[(cse_var_2 + 199)] = 0f32
- compute_5[(cse_var_2 + 200)] = 0f32
- compute_5[(cse_var_2 + 201)] = 0f32
- compute_5[(cse_var_2 + 202)] = 0f32
- compute_5[(cse_var_2 + 203)] = 0f32
- compute_5[(cse_var_2 + 204)] = 0f32
- compute_5[(cse_var_2 + 205)] = 0f32
- compute_5[(cse_var_2 + 206)] = 0f32
- compute_5[(cse_var_2 + 207)] = 0f32
- compute_5[(cse_var_2 + 224)] = 0f32
- compute_5[(cse_var_2 + 225)] = 0f32
- compute_5[(cse_var_2 + 226)] = 0f32
- compute_5[(cse_var_2 + 227)] = 0f32
- compute_5[(cse_var_2 + 228)] = 0f32
- compute_5[(cse_var_2 + 229)] = 0f32
- compute_5[(cse_var_2 + 230)] = 0f32
- compute_5[(cse_var_2 + 231)] = 0f32
- compute_5[(cse_var_2 + 232)] = 0f32
- compute_5[(cse_var_2 + 233)] = 0f32
- compute_5[(cse_var_2 + 234)] = 0f32
- compute_5[(cse_var_2 + 235)] = 0f32
- compute_5[(cse_var_2 + 236)] = 0f32
- compute_5[(cse_var_2 + 237)] = 0f32
- compute_5[(cse_var_2 + 238)] = 0f32
- compute_5[(cse_var_2 + 239)] = 0f32
- compute_5[(cse_var_2 + 256)] = 0f32
- compute_5[(cse_var_2 + 257)] = 0f32
- compute_5[(cse_var_2 + 258)] = 0f32
- compute_5[(cse_var_2 + 259)] = 0f32
- compute_5[(cse_var_2 + 260)] = 0f32
- compute_5[(cse_var_2 + 261)] = 0f32
- compute_5[(cse_var_2 + 262)] = 0f32
- compute_5[(cse_var_2 + 263)] = 0f32
- compute_5[(cse_var_2 + 264)] = 0f32
- compute_5[(cse_var_2 + 265)] = 0f32
- compute_5[(cse_var_2 + 266)] = 0f32
- compute_5[(cse_var_2 + 267)] = 0f32
- compute_5[(cse_var_2 + 268)] = 0f32
- compute_5[(cse_var_2 + 269)] = 0f32
- compute_5[(cse_var_2 + 270)] = 0f32
- compute_5[(cse_var_2 + 271)] = 0f32
- compute_5[(cse_var_2 + 288)] = 0f32
- compute_5[(cse_var_2 + 289)] = 0f32
- compute_5[(cse_var_2 + 290)] = 0f32
- compute_5[(cse_var_2 + 291)] = 0f32
- compute_5[(cse_var_2 + 292)] = 0f32
- compute_5[(cse_var_2 + 293)] = 0f32
- compute_5[(cse_var_2 + 294)] = 0f32
- compute_5[(cse_var_2 + 295)] = 0f32
- compute_5[(cse_var_2 + 296)] = 0f32
- compute_5[(cse_var_2 + 297)] = 0f32
- compute_5[(cse_var_2 + 298)] = 0f32
- compute_5[(cse_var_2 + 299)] = 0f32
- compute_5[(cse_var_2 + 300)] = 0f32
- compute_5[(cse_var_2 + 301)] = 0f32
- compute_5[(cse_var_2 + 302)] = 0f32
- compute_5[(cse_var_2 + 303)] = 0f32
- compute_5[(cse_var_2 + 320)] = 0f32
- compute_5[(cse_var_2 + 321)] = 0f32
- compute_5[(cse_var_2 + 322)] = 0f32
- compute_5[(cse_var_2 + 323)] = 0f32
- compute_5[(cse_var_2 + 324)] = 0f32
- compute_5[(cse_var_2 + 325)] = 0f32
- compute_5[(cse_var_2 + 326)] = 0f32
- compute_5[(cse_var_2 + 327)] = 0f32
- compute_5[(cse_var_2 + 328)] = 0f32
- compute_5[(cse_var_2 + 329)] = 0f32
- compute_5[(cse_var_2 + 330)] = 0f32
- compute_5[(cse_var_2 + 331)] = 0f32
- compute_5[(cse_var_2 + 332)] = 0f32
- compute_5[(cse_var_2 + 333)] = 0f32
- compute_5[(cse_var_2 + 334)] = 0f32
- compute_5[(cse_var_2 + 335)] = 0f32
- compute_5[(cse_var_2 + 352)] = 0f32
- compute_5[(cse_var_2 + 353)] = 0f32
- compute_5[(cse_var_2 + 354)] = 0f32
- compute_5[(cse_var_2 + 355)] = 0f32
- compute_5[(cse_var_2 + 356)] = 0f32
- compute_5[(cse_var_2 + 357)] = 0f32
- compute_5[(cse_var_2 + 358)] = 0f32
- compute_5[(cse_var_2 + 359)] = 0f32
- compute_5[(cse_var_2 + 360)] = 0f32
- compute_5[(cse_var_2 + 361)] = 0f32
- compute_5[(cse_var_2 + 362)] = 0f32
- compute_5[(cse_var_2 + 363)] = 0f32
- compute_5[(cse_var_2 + 364)] = 0f32
- compute_5[(cse_var_2 + 365)] = 0f32
- compute_5[(cse_var_2 + 366)] = 0f32
- compute_5[(cse_var_2 + 367)] = 0f32
- compute_5[(cse_var_2 + 384)] = 0f32
- compute_5[(cse_var_2 + 385)] = 0f32
- compute_5[(cse_var_2 + 386)] = 0f32
- compute_5[(cse_var_2 + 387)] = 0f32
- compute_5[(cse_var_2 + 388)] = 0f32
- compute_5[(cse_var_2 + 389)] = 0f32
- compute_5[(cse_var_2 + 390)] = 0f32
- compute_5[(cse_var_2 + 391)] = 0f32
- compute_5[(cse_var_2 + 392)] = 0f32
- compute_5[(cse_var_2 + 393)] = 0f32
- compute_5[(cse_var_2 + 394)] = 0f32
- compute_5[(cse_var_2 + 395)] = 0f32
- compute_5[(cse_var_2 + 396)] = 0f32
- compute_5[(cse_var_2 + 397)] = 0f32
- compute_5[(cse_var_2 + 398)] = 0f32
- compute_5[(cse_var_2 + 399)] = 0f32
- compute_5[(cse_var_2 + 416)] = 0f32
- compute_5[(cse_var_2 + 417)] = 0f32
- compute_5[(cse_var_2 + 418)] = 0f32
- compute_5[(cse_var_2 + 419)] = 0f32
- compute_5[(cse_var_2 + 420)] = 0f32
- compute_5[(cse_var_2 + 421)] = 0f32
- compute_5[(cse_var_2 + 422)] = 0f32
- compute_5[(cse_var_2 + 423)] = 0f32
- compute_5[(cse_var_2 + 424)] = 0f32
- compute_5[(cse_var_2 + 425)] = 0f32
- compute_5[(cse_var_2 + 426)] = 0f32
- compute_5[(cse_var_2 + 427)] = 0f32
- compute_5[(cse_var_2 + 428)] = 0f32
- compute_5[(cse_var_2 + 429)] = 0f32
- compute_5[(cse_var_2 + 430)] = 0f32
- compute_5[(cse_var_2 + 431)] = 0f32
- compute_5[(cse_var_2 + 448)] = 0f32
- compute_5[(cse_var_2 + 449)] = 0f32
- compute_5[(cse_var_2 + 450)] = 0f32
- compute_5[(cse_var_2 + 451)] = 0f32
- compute_5[(cse_var_2 + 452)] = 0f32
- compute_5[(cse_var_2 + 453)] = 0f32
- compute_5[(cse_var_2 + 454)] = 0f32
- compute_5[(cse_var_2 + 455)] = 0f32
- compute_5[(cse_var_2 + 456)] = 0f32
- compute_5[(cse_var_2 + 457)] = 0f32
- compute_5[(cse_var_2 + 458)] = 0f32
- compute_5[(cse_var_2 + 459)] = 0f32
- compute_5[(cse_var_2 + 460)] = 0f32
- compute_5[(cse_var_2 + 461)] = 0f32
- compute_5[(cse_var_2 + 462)] = 0f32
- compute_5[(cse_var_2 + 463)] = 0f32
- compute_5[(cse_var_2 + 480)] = 0f32
- compute_5[(cse_var_2 + 481)] = 0f32
- compute_5[(cse_var_2 + 482)] = 0f32
- compute_5[(cse_var_2 + 483)] = 0f32
- compute_5[(cse_var_2 + 484)] = 0f32
- compute_5[(cse_var_2 + 485)] = 0f32
- compute_5[(cse_var_2 + 486)] = 0f32
- compute_5[(cse_var_2 + 487)] = 0f32
- compute_5[(cse_var_2 + 488)] = 0f32
- compute_5[(cse_var_2 + 489)] = 0f32
- compute_5[(cse_var_2 + 490)] = 0f32
- compute_5[(cse_var_2 + 491)] = 0f32
- compute_5[(cse_var_2 + 492)] = 0f32
- compute_5[(cse_var_2 + 493)] = 0f32
- compute_5[(cse_var_2 + 494)] = 0f32
- compute_5[(cse_var_2 + 495)] = 0f32
- for (elem_idx: int32, 0, (placeholder_3[(cse_var_1 + 1)] - placeholder_3[cse_var_1])) {
- let cse_var_259: int32 = (i0.outer*4096)
- let cse_var_258: int32 = (elem_idx*16)
- let cse_var_257: int32 = (cse_var_2 + 99)
- let cse_var_256: int32 = (cse_var_2 + 98)
- let cse_var_255: int32 = (cse_var_2 + 97)
- let cse_var_254: int32 = (cse_var_2 + 96)
- let cse_var_253: int32 = (cse_var_2 + 9)
- let cse_var_252: int32 = (cse_var_2 + 8)
- let cse_var_251: int32 = (cse_var_2 + 79)
- let cse_var_250: int32 = (cse_var_2 + 78)
- let cse_var_249: int32 = (cse_var_2 + 77)
- let cse_var_248: int32 = (cse_var_2 + 76)
- let cse_var_247: int32 = (cse_var_2 + 75)
- let cse_var_246: int32 = (cse_var_2 + 74)
- let cse_var_245: int32 = (cse_var_2 + 73)
- let cse_var_244: int32 = (cse_var_2 + 72)
- let cse_var_243: int32 = (cse_var_2 + 71)
- let cse_var_242: int32 = (cse_var_2 + 70)
- let cse_var_241: int32 = (cse_var_2 + 7)
- let cse_var_240: int32 = (cse_var_2 + 69)
- let cse_var_239: int32 = (cse_var_2 + 68)
- let cse_var_238: int32 = (cse_var_2 + 67)
- let cse_var_237: int32 = (cse_var_2 + 66)
- let cse_var_236: int32 = (cse_var_2 + 65)
- let cse_var_235: int32 = (cse_var_2 + 64)
- let cse_var_234: int32 = (cse_var_2 + 6)
- let cse_var_233: int32 = (cse_var_2 + 5)
- let cse_var_232: int32 = (cse_var_2 + 495)
- let cse_var_231: int32 = (cse_var_2 + 494)
- let cse_var_230: int32 = (cse_var_2 + 493)
- let cse_var_229: int32 = (cse_var_2 + 492)
- let cse_var_228: int32 = (cse_var_2 + 491)
- let cse_var_227: int32 = (cse_var_2 + 490)
- let cse_var_226: int32 = (cse_var_2 + 489)
- let cse_var_225: int32 = (cse_var_2 + 488)
- let cse_var_224: int32 = (cse_var_2 + 487)
- let cse_var_223: int32 = (cse_var_2 + 486)
- let cse_var_222: int32 = (cse_var_2 + 485)
- let cse_var_221: int32 = (cse_var_2 + 484)
- let cse_var_220: int32 = (cse_var_2 + 483)
- let cse_var_219: int32 = (cse_var_2 + 482)
- let cse_var_218: int32 = (cse_var_2 + 481)
- let cse_var_217: int32 = (cse_var_2 + 480)
- let cse_var_216: int32 = (cse_var_2 + 47)
- let cse_var_215: int32 = (cse_var_2 + 463)
- let cse_var_214: int32 = (cse_var_2 + 462)
- let cse_var_213: int32 = (cse_var_2 + 461)
- let cse_var_212: int32 = (cse_var_2 + 460)
- let cse_var_211: int32 = (cse_var_2 + 46)
- let cse_var_210: int32 = (cse_var_2 + 459)
- let cse_var_209: int32 = (cse_var_2 + 458)
- let cse_var_208: int32 = (cse_var_2 + 457)
- let cse_var_207: int32 = (cse_var_2 + 456)
- let cse_var_206: int32 = (cse_var_2 + 455)
- let cse_var_205: int32 = (cse_var_2 + 454)
- let cse_var_204: int32 = (cse_var_2 + 453)
- let cse_var_203: int32 = (cse_var_2 + 452)
- let cse_var_202: int32 = (cse_var_2 + 451)
- let cse_var_201: int32 = (cse_var_2 + 450)
- let cse_var_200: int32 = (cse_var_2 + 45)
- let cse_var_199: int32 = (cse_var_2 + 449)
- let cse_var_198: int32 = (cse_var_2 + 448)
- let cse_var_197: int32 = (cse_var_2 + 44)
- let cse_var_196: int32 = (cse_var_2 + 431)
- let cse_var_195: int32 = (cse_var_2 + 430)
- let cse_var_194: int32 = (cse_var_2 + 43)
- let cse_var_193: int32 = (cse_var_2 + 429)
- let cse_var_192: int32 = (cse_var_2 + 428)
- let cse_var_191: int32 = (cse_var_2 + 427)
- let cse_var_190: int32 = (cse_var_2 + 426)
- let cse_var_189: int32 = (cse_var_2 + 425)
- let cse_var_188: int32 = (cse_var_2 + 424)
- let cse_var_187: int32 = (cse_var_2 + 423)
- let cse_var_186: int32 = (cse_var_2 + 422)
- let cse_var_185: int32 = (cse_var_2 + 421)
- let cse_var_184: int32 = (cse_var_2 + 420)
- let cse_var_183: int32 = (cse_var_2 + 42)
- let cse_var_182: int32 = (cse_var_2 + 419)
- let cse_var_181: int32 = (cse_var_2 + 418)
- let cse_var_180: int32 = (cse_var_2 + 417)
- let cse_var_179: int32 = (cse_var_2 + 416)
- let cse_var_178: int32 = (cse_var_2 + 41)
- let cse_var_177: int32 = (cse_var_2 + 40)
- let cse_var_176: int32 = (cse_var_2 + 4)
- let cse_var_175: int32 = (cse_var_2 + 399)
- let cse_var_174: int32 = (cse_var_2 + 398)
- let cse_var_173: int32 = (cse_var_2 + 397)
- let cse_var_172: int32 = (cse_var_2 + 396)
- let cse_var_171: int32 = (cse_var_2 + 395)
- let cse_var_170: int32 = (cse_var_2 + 394)
- let cse_var_169: int32 = (cse_var_2 + 393)
- let cse_var_168: int32 = (cse_var_2 + 392)
- let cse_var_167: int32 = (cse_var_2 + 391)
- let cse_var_166: int32 = (cse_var_2 + 390)
- let cse_var_165: int32 = (cse_var_2 + 39)
- let cse_var_164: int32 = (cse_var_2 + 389)
- let cse_var_163: int32 = (cse_var_2 + 388)
- let cse_var_162: int32 = (cse_var_2 + 387)
- let cse_var_161: int32 = (cse_var_2 + 386)
- let cse_var_160: int32 = (cse_var_2 + 385)
- let cse_var_159: int32 = (cse_var_2 + 384)
- let cse_var_158: int32 = (cse_var_2 + 38)
- let cse_var_157: int32 = (cse_var_2 + 37)
- let cse_var_156: int32 = (cse_var_2 + 367)
- let cse_var_155: int32 = (cse_var_2 + 366)
- let cse_var_154: int32 = (cse_var_2 + 365)
- let cse_var_153: int32 = (cse_var_2 + 364)
- let cse_var_152: int32 = (cse_var_2 + 363)
- let cse_var_151: int32 = (cse_var_2 + 362)
- let cse_var_150: int32 = (cse_var_2 + 361)
- let cse_var_149: int32 = (cse_var_2 + 360)
- let cse_var_148: int32 = (cse_var_2 + 36)
- let cse_var_147: int32 = (cse_var_2 + 359)
- let cse_var_146: int32 = (cse_var_2 + 358)
- let cse_var_145: int32 = (cse_var_2 + 357)
- let cse_var_144: int32 = (cse_var_2 + 356)
- let cse_var_143: int32 = (cse_var_2 + 355)
- let cse_var_142: int32 = (cse_var_2 + 354)
- let cse_var_141: int32 = (cse_var_2 + 353)
- let cse_var_140: int32 = (cse_var_2 + 352)
- let cse_var_139: int32 = (cse_var_2 + 35)
- let cse_var_138: int32 = (cse_var_2 + 34)
- let cse_var_137: int32 = (cse_var_2 + 335)
- let cse_var_136: int32 = (cse_var_2 + 334)
- let cse_var_135: int32 = (cse_var_2 + 333)
- let cse_var_134: int32 = (cse_var_2 + 332)
- let cse_var_133: int32 = (cse_var_2 + 331)
- let cse_var_132: int32 = (cse_var_2 + 330)
- let cse_var_131: int32 = (cse_var_2 + 33)
- let cse_var_130: int32 = (cse_var_2 + 329)
- let cse_var_129: int32 = (cse_var_2 + 328)
- let cse_var_128: int32 = (cse_var_2 + 327)
- let cse_var_127: int32 = (cse_var_2 + 326)
- let cse_var_126: int32 = (cse_var_2 + 325)
- let cse_var_125: int32 = (cse_var_2 + 324)
- let cse_var_124: int32 = (cse_var_2 + 323)
- let cse_var_123: int32 = (cse_var_2 + 322)
- let cse_var_122: int32 = (cse_var_2 + 321)
- let cse_var_121: int32 = (cse_var_2 + 320)
- let cse_var_120: int32 = (cse_var_2 + 32)
- let cse_var_119: int32 = (cse_var_2 + 303)
- let cse_var_118: int32 = (cse_var_2 + 302)
- let cse_var_117: int32 = (cse_var_2 + 301)
- let cse_var_116: int32 = (cse_var_2 + 300)
- let cse_var_115: int32 = (cse_var_2 + 3)
- let cse_var_114: int32 = (cse_var_2 + 299)
- let cse_var_113: int32 = (cse_var_2 + 298)
- let cse_var_112: int32 = (cse_var_2 + 297)
- let cse_var_111: int32 = (cse_var_2 + 296)
- let cse_var_110: int32 = (cse_var_2 + 295)
- let cse_var_109: int32 = (cse_var_2 + 294)
- let cse_var_108: int32 = (cse_var_2 + 293)
- let cse_var_107: int32 = (cse_var_2 + 292)
- let cse_var_106: int32 = (cse_var_2 + 291)
- let cse_var_105: int32 = (cse_var_2 + 290)
- let cse_var_104: int32 = (cse_var_2 + 289)
- let cse_var_103: int32 = (cse_var_2 + 288)
- let cse_var_102: int32 = (cse_var_2 + 271)
- let cse_var_101: int32 = (cse_var_2 + 270)
- let cse_var_100: int32 = (cse_var_2 + 269)
- let cse_var_99: int32 = (cse_var_2 + 268)
- let cse_var_98: int32 = (cse_var_2 + 267)
- let cse_var_97: int32 = (cse_var_2 + 266)
- let cse_var_96: int32 = (cse_var_2 + 265)
- let cse_var_95: int32 = (cse_var_2 + 264)
- let cse_var_94: int32 = (cse_var_2 + 263)
- let cse_var_93: int32 = (cse_var_2 + 262)
- let cse_var_92: int32 = (cse_var_2 + 261)
- let cse_var_91: int32 = (cse_var_2 + 260)
- let cse_var_90: int32 = (cse_var_2 + 259)
- let cse_var_89: int32 = (cse_var_2 + 258)
- let cse_var_88: int32 = (cse_var_2 + 257)
- let cse_var_87: int32 = (cse_var_2 + 256)
- let cse_var_86: int32 = (cse_var_2 + 239)
- let cse_var_85: int32 = (cse_var_2 + 238)
- let cse_var_84: int32 = (cse_var_2 + 237)
- let cse_var_83: int32 = (cse_var_2 + 236)
- let cse_var_82: int32 = (cse_var_2 + 235)
- let cse_var_81: int32 = (cse_var_2 + 234)
- let cse_var_80: int32 = (cse_var_2 + 233)
- let cse_var_79: int32 = (cse_var_2 + 232)
- let cse_var_78: int32 = (cse_var_2 + 231)
- let cse_var_77: int32 = (cse_var_2 + 230)
- let cse_var_76: int32 = (cse_var_2 + 229)
- let cse_var_75: int32 = (cse_var_2 + 228)
- let cse_var_74: int32 = (cse_var_2 + 227)
- let cse_var_73: int32 = (cse_var_2 + 226)
- let cse_var_72: int32 = (cse_var_2 + 225)
- let cse_var_71: int32 = (cse_var_2 + 224)
- let cse_var_70: int32 = (cse_var_2 + 207)
- let cse_var_69: int32 = (cse_var_2 + 206)
- let cse_var_68: int32 = (cse_var_2 + 205)
- let cse_var_67: int32 = (cse_var_2 + 204)
- let cse_var_66: int32 = (cse_var_2 + 203)
- let cse_var_65: int32 = (cse_var_2 + 202)
- let cse_var_64: int32 = (cse_var_2 + 201)
- let cse_var_63: int32 = (cse_var_2 + 200)
- let cse_var_62: int32 = (cse_var_2 + 2)
- let cse_var_61: int32 = (cse_var_2 + 199)
- let cse_var_60: int32 = (cse_var_2 + 198)
- let cse_var_59: int32 = (cse_var_2 + 197)
- let cse_var_58: int32 = (cse_var_2 + 196)
- let cse_var_57: int32 = (cse_var_2 + 195)
- let cse_var_56: int32 = (cse_var_2 + 194)
- let cse_var_55: int32 = (cse_var_2 + 193)
- let cse_var_54: int32 = (cse_var_2 + 192)
- let cse_var_53: int32 = (cse_var_2 + 175)
- let cse_var_52: int32 = (cse_var_2 + 174)
- let cse_var_51: int32 = (cse_var_2 + 173)
- let cse_var_50: int32 = (cse_var_2 + 172)
- let cse_var_49: int32 = (cse_var_2 + 171)
- let cse_var_48: int32 = (cse_var_2 + 170)
- let cse_var_47: int32 = (cse_var_2 + 169)
- let cse_var_46: int32 = (cse_var_2 + 168)
- let cse_var_45: int32 = (cse_var_2 + 167)
- let cse_var_44: int32 = (cse_var_2 + 166)
- let cse_var_43: int32 = (cse_var_2 + 165)
- let cse_var_42: int32 = (cse_var_2 + 164)
- let cse_var_41: int32 = (cse_var_2 + 163)
- let cse_var_40: int32 = (cse_var_2 + 162)
- let cse_var_39: int32 = (cse_var_2 + 161)
- let cse_var_38: int32 = (cse_var_2 + 160)
- let cse_var_37: int32 = (cse_var_2 + 15)
- let cse_var_36: int32 = (cse_var_2 + 143)
- let cse_var_35: int32 = (cse_var_2 + 142)
- let cse_var_34: int32 = (cse_var_2 + 141)
- let cse_var_33: int32 = (cse_var_2 + 140)
- let cse_var_32: int32 = (cse_var_2 + 14)
- let cse_var_31: int32 = (cse_var_2 + 139)
- let cse_var_30: int32 = (cse_var_2 + 138)
- let cse_var_29: int32 = (cse_var_2 + 137)
- let cse_var_28: int32 = (cse_var_2 + 136)
- let cse_var_27: int32 = (cse_var_2 + 135)
- let cse_var_26: int32 = (cse_var_2 + 134)
- let cse_var_25: int32 = (cse_var_2 + 133)
- let cse_var_24: int32 = (cse_var_2 + 132)
- let cse_var_23: int32 = (cse_var_2 + 131)
- let cse_var_22: int32 = (cse_var_2 + 130)
- let cse_var_21: int32 = (cse_var_2 + 13)
- let cse_var_20: int32 = (cse_var_2 + 129)
- let cse_var_19: int32 = (cse_var_2 + 128)
- let cse_var_18: int32 = (cse_var_2 + 12)
- let cse_var_17: int32 = (cse_var_2 + 111)
- let cse_var_16: int32 = (cse_var_2 + 110)
- let cse_var_15: int32 = (cse_var_2 + 11)
- let cse_var_14: int32 = (cse_var_2 + 109)
- let cse_var_13: int32 = (cse_var_2 + 108)
- let cse_var_12: int32 = (cse_var_2 + 107)
- let cse_var_11: int32 = (cse_var_2 + 106)
- let cse_var_10: int32 = (cse_var_2 + 105)
- let cse_var_9: int32 = (cse_var_2 + 104)
- let cse_var_8: int32 = (cse_var_2 + 103)
- let cse_var_7: int32 = (cse_var_2 + 102)
- let cse_var_6: int32 = (cse_var_2 + 101)
- let cse_var_5: int32 = (cse_var_2 + 100)
- let cse_var_4: int32 = (cse_var_2 + 10)
- let cse_var_3: int32 = (cse_var_2 + 1)
- {
- compute_5[cse_var_2] = (compute_5[cse_var_2] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_3] = (compute_5[cse_var_3] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_62] = (compute_5[cse_var_62] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_115] = (compute_5[cse_var_115] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_176] = (compute_5[cse_var_176] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_233] = (compute_5[cse_var_233] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_234] = (compute_5[cse_var_234] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_241] = (compute_5[cse_var_241] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_252] = (compute_5[cse_var_252] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_253] = (compute_5[cse_var_253] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_4] = (compute_5[cse_var_4] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_21] = (compute_5[cse_var_21] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_32] = (compute_5[cse_var_32] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_37] = (compute_5[cse_var_37] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_120] = (compute_5[cse_var_120] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_131] = (compute_5[cse_var_131] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_138] = (compute_5[cse_var_138] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_139] = (compute_5[cse_var_139] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_148] = (compute_5[cse_var_148] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_157] = (compute_5[cse_var_157] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_158] = (compute_5[cse_var_158] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_165] = (compute_5[cse_var_165] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_177] = (compute_5[cse_var_177] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_178] = (compute_5[cse_var_178] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_183] = (compute_5[cse_var_183] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_194] = (compute_5[cse_var_194] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_197] = (compute_5[cse_var_197] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_200] = (compute_5[cse_var_200] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_211] = (compute_5[cse_var_211] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_216] = (compute_5[cse_var_216] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_235] = (compute_5[cse_var_235] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_236] = (compute_5[cse_var_236] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_237] = (compute_5[cse_var_237] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_238] = (compute_5[cse_var_238] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_239] = (compute_5[cse_var_239] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_240] = (compute_5[cse_var_240] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_242] = (compute_5[cse_var_242] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_243] = (compute_5[cse_var_243] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_244] = (compute_5[cse_var_244] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_245] = (compute_5[cse_var_245] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_246] = (compute_5[cse_var_246] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_247] = (compute_5[cse_var_247] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_248] = (compute_5[cse_var_248] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_249] = (compute_5[cse_var_249] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_250] = (compute_5[cse_var_250] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_251] = (compute_5[cse_var_251] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_254] = (compute_5[cse_var_254] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_255] = (compute_5[cse_var_255] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_256] = (compute_5[cse_var_256] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_257] = (compute_5[cse_var_257] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_19] = (compute_5[cse_var_19] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_20] = (compute_5[cse_var_20] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_22] = (compute_5[cse_var_22] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_23] = (compute_5[cse_var_23] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_24] = (compute_5[cse_var_24] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_25] = (compute_5[cse_var_25] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_26] = (compute_5[cse_var_26] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_27] = (compute_5[cse_var_27] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_28] = (compute_5[cse_var_28] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_29] = (compute_5[cse_var_29] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_30] = (compute_5[cse_var_30] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_31] = (compute_5[cse_var_31] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_33] = (compute_5[cse_var_33] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_34] = (compute_5[cse_var_34] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_35] = (compute_5[cse_var_35] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_36] = (compute_5[cse_var_36] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_38] = (compute_5[cse_var_38] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_39] = (compute_5[cse_var_39] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_40] = (compute_5[cse_var_40] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_41] = (compute_5[cse_var_41] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_42] = (compute_5[cse_var_42] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_43] = (compute_5[cse_var_43] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_44] = (compute_5[cse_var_44] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_45] = (compute_5[cse_var_45] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_46] = (compute_5[cse_var_46] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_47] = (compute_5[cse_var_47] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_48] = (compute_5[cse_var_48] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_49] = (compute_5[cse_var_49] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_50] = (compute_5[cse_var_50] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_51] = (compute_5[cse_var_51] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_52] = (compute_5[cse_var_52] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_53] = (compute_5[cse_var_53] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_54] = (compute_5[cse_var_54] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_55] = (compute_5[cse_var_55] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_56] = (compute_5[cse_var_56] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_57] = (compute_5[cse_var_57] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_58] = (compute_5[cse_var_58] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_59] = (compute_5[cse_var_59] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_60] = (compute_5[cse_var_60] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_61] = (compute_5[cse_var_61] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_63] = (compute_5[cse_var_63] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_64] = (compute_5[cse_var_64] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_65] = (compute_5[cse_var_65] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_66] = (compute_5[cse_var_66] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_67] = (compute_5[cse_var_67] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_68] = (compute_5[cse_var_68] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_69] = (compute_5[cse_var_69] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_70] = (compute_5[cse_var_70] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_71] = (compute_5[cse_var_71] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_72] = (compute_5[cse_var_72] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_73] = (compute_5[cse_var_73] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_74] = (compute_5[cse_var_74] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_75] = (compute_5[cse_var_75] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_76] = (compute_5[cse_var_76] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_77] = (compute_5[cse_var_77] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_78] = (compute_5[cse_var_78] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_79] = (compute_5[cse_var_79] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_80] = (compute_5[cse_var_80] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_81] = (compute_5[cse_var_81] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_82] = (compute_5[cse_var_82] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_83] = (compute_5[cse_var_83] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_84] = (compute_5[cse_var_84] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_85] = (compute_5[cse_var_85] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_86] = (compute_5[cse_var_86] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_87] = (compute_5[cse_var_87] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_88] = (compute_5[cse_var_88] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_89] = (compute_5[cse_var_89] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_90] = (compute_5[cse_var_90] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_91] = (compute_5[cse_var_91] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_92] = (compute_5[cse_var_92] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_93] = (compute_5[cse_var_93] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_94] = (compute_5[cse_var_94] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_95] = (compute_5[cse_var_95] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_96] = (compute_5[cse_var_96] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_97] = (compute_5[cse_var_97] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_98] = (compute_5[cse_var_98] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_99] = (compute_5[cse_var_99] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_100] = (compute_5[cse_var_100] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_101] = (compute_5[cse_var_101] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_102] = (compute_5[cse_var_102] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_103] = (compute_5[cse_var_103] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_104] = (compute_5[cse_var_104] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_105] = (compute_5[cse_var_105] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_106] = (compute_5[cse_var_106] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_107] = (compute_5[cse_var_107] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_108] = (compute_5[cse_var_108] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_109] = (compute_5[cse_var_109] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_110] = (compute_5[cse_var_110] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_111] = (compute_5[cse_var_111] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_112] = (compute_5[cse_var_112] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_113] = (compute_5[cse_var_113] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_114] = (compute_5[cse_var_114] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_116] = (compute_5[cse_var_116] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_117] = (compute_5[cse_var_117] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_118] = (compute_5[cse_var_118] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_119] = (compute_5[cse_var_119] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_121] = (compute_5[cse_var_121] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_122] = (compute_5[cse_var_122] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_123] = (compute_5[cse_var_123] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_124] = (compute_5[cse_var_124] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_125] = (compute_5[cse_var_125] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_126] = (compute_5[cse_var_126] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_127] = (compute_5[cse_var_127] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_128] = (compute_5[cse_var_128] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_129] = (compute_5[cse_var_129] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_130] = (compute_5[cse_var_130] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_132] = (compute_5[cse_var_132] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_133] = (compute_5[cse_var_133] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_134] = (compute_5[cse_var_134] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_135] = (compute_5[cse_var_135] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_136] = (compute_5[cse_var_136] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_137] = (compute_5[cse_var_137] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_140] = (compute_5[cse_var_140] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_141] = (compute_5[cse_var_141] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_142] = (compute_5[cse_var_142] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_143] = (compute_5[cse_var_143] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_144] = (compute_5[cse_var_144] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_145] = (compute_5[cse_var_145] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_146] = (compute_5[cse_var_146] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_147] = (compute_5[cse_var_147] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_149] = (compute_5[cse_var_149] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_150] = (compute_5[cse_var_150] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_151] = (compute_5[cse_var_151] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_152] = (compute_5[cse_var_152] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_153] = (compute_5[cse_var_153] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_154] = (compute_5[cse_var_154] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_155] = (compute_5[cse_var_155] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_156] = (compute_5[cse_var_156] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_159] = (compute_5[cse_var_159] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_160] = (compute_5[cse_var_160] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_161] = (compute_5[cse_var_161] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_162] = (compute_5[cse_var_162] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_163] = (compute_5[cse_var_163] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_164] = (compute_5[cse_var_164] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_166] = (compute_5[cse_var_166] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_167] = (compute_5[cse_var_167] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_168] = (compute_5[cse_var_168] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_169] = (compute_5[cse_var_169] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_170] = (compute_5[cse_var_170] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_171] = (compute_5[cse_var_171] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_172] = (compute_5[cse_var_172] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_173] = (compute_5[cse_var_173] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_174] = (compute_5[cse_var_174] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_175] = (compute_5[cse_var_175] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_179] = (compute_5[cse_var_179] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_180] = (compute_5[cse_var_180] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_181] = (compute_5[cse_var_181] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_182] = (compute_5[cse_var_182] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_184] = (compute_5[cse_var_184] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_185] = (compute_5[cse_var_185] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_186] = (compute_5[cse_var_186] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_187] = (compute_5[cse_var_187] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_188] = (compute_5[cse_var_188] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_189] = (compute_5[cse_var_189] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_190] = (compute_5[cse_var_190] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_191] = (compute_5[cse_var_191] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_192] = (compute_5[cse_var_192] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_193] = (compute_5[cse_var_193] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_195] = (compute_5[cse_var_195] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_196] = (compute_5[cse_var_196] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_198] = (compute_5[cse_var_198] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_199] = (compute_5[cse_var_199] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_201] = (compute_5[cse_var_201] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_202] = (compute_5[cse_var_202] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_203] = (compute_5[cse_var_203] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_204] = (compute_5[cse_var_204] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_205] = (compute_5[cse_var_205] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_206] = (compute_5[cse_var_206] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_207] = (compute_5[cse_var_207] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_208] = (compute_5[cse_var_208] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_209] = (compute_5[cse_var_209] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_210] = (compute_5[cse_var_210] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_212] = (compute_5[cse_var_212] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_213] = (compute_5[cse_var_213] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_214] = (compute_5[cse_var_214] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_215] = (compute_5[cse_var_215] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_217] = (compute_5[cse_var_217] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_218] = (compute_5[cse_var_218] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_219] = (compute_5[cse_var_219] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_220] = (compute_5[cse_var_220] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_221] = (compute_5[cse_var_221] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_222] = (compute_5[cse_var_222] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_223] = (compute_5[cse_var_223] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_224] = (compute_5[cse_var_224] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_225] = (compute_5[cse_var_225] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_226] = (compute_5[cse_var_226] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_227] = (compute_5[cse_var_227] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_228] = (compute_5[cse_var_228] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_229] = (compute_5[cse_var_229] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_230] = (compute_5[cse_var_230] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_231] = (compute_5[cse_var_231] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_232] = (compute_5[cse_var_232] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
+ preflattened_buffer_map = {placeholder_6: placeholder_15: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_8: placeholder_16: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_17: Buffer(placeholder_12, int32, [4916], []), placeholder_9: placeholder_18: Buffer(placeholder_14, float32, [128, 512], []), placeholder_5: placeholder_19: Buffer(placeholder_10, float32, [128, 256], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], [])} {
+ for (i0.outer.i1.outer.fused: int32, 0, 16) "parallel" {
+ allocate(compute_4: Pointer(global float32), float32, [4096]), storage_scope = global {
+ for (i.outer.inner: int32, 0, 16) {
+ for (nb_j.inner: int32, 0, 2) {
+ let cse_var_2: int32 = ((i0.outer.i1.outer.fused*2) + nb_j.inner)
+ let cse_var_1: int32 = ((i.outer.inner*256) + (nb_j.inner*16))
+ {
+ compute_5: Buffer(compute_4, float32, [4096], [])[cse_var_1] = 0f32
+ compute_5[(cse_var_1 + 1)] = 0f32
+ compute_5[(cse_var_1 + 2)] = 0f32
+ compute_5[(cse_var_1 + 3)] = 0f32
+ compute_5[(cse_var_1 + 4)] = 0f32
+ compute_5[(cse_var_1 + 5)] = 0f32
+ compute_5[(cse_var_1 + 6)] = 0f32
+ compute_5[(cse_var_1 + 7)] = 0f32
+ compute_5[(cse_var_1 + 8)] = 0f32
+ compute_5[(cse_var_1 + 9)] = 0f32
+ compute_5[(cse_var_1 + 10)] = 0f32
+ compute_5[(cse_var_1 + 11)] = 0f32
+ compute_5[(cse_var_1 + 12)] = 0f32
+ compute_5[(cse_var_1 + 13)] = 0f32
+ compute_5[(cse_var_1 + 14)] = 0f32
+ compute_5[(cse_var_1 + 15)] = 0f32
+ compute_5[(cse_var_1 + 32)] = 0f32
+ compute_5[(cse_var_1 + 33)] = 0f32
+ compute_5[(cse_var_1 + 34)] = 0f32
+ compute_5[(cse_var_1 + 35)] = 0f32
+ compute_5[(cse_var_1 + 36)] = 0f32
+ compute_5[(cse_var_1 + 37)] = 0f32
+ compute_5[(cse_var_1 + 38)] = 0f32
+ compute_5[(cse_var_1 + 39)] = 0f32
+ compute_5[(cse_var_1 + 40)] = 0f32
+ compute_5[(cse_var_1 + 41)] = 0f32
+ compute_5[(cse_var_1 + 42)] = 0f32
+ compute_5[(cse_var_1 + 43)] = 0f32
+ compute_5[(cse_var_1 + 44)] = 0f32
+ compute_5[(cse_var_1 + 45)] = 0f32
+ compute_5[(cse_var_1 + 46)] = 0f32
+ compute_5[(cse_var_1 + 47)] = 0f32
+ compute_5[(cse_var_1 + 64)] = 0f32
+ compute_5[(cse_var_1 + 65)] = 0f32
+ compute_5[(cse_var_1 + 66)] = 0f32
+ compute_5[(cse_var_1 + 67)] = 0f32
+ compute_5[(cse_var_1 + 68)] = 0f32
+ compute_5[(cse_var_1 + 69)] = 0f32
+ compute_5[(cse_var_1 + 70)] = 0f32
+ compute_5[(cse_var_1 + 71)] = 0f32
+ compute_5[(cse_var_1 + 72)] = 0f32
+ compute_5[(cse_var_1 + 73)] = 0f32
+ compute_5[(cse_var_1 + 74)] = 0f32
+ compute_5[(cse_var_1 + 75)] = 0f32
+ compute_5[(cse_var_1 + 76)] = 0f32
+ compute_5[(cse_var_1 + 77)] = 0f32
+ compute_5[(cse_var_1 + 78)] = 0f32
+ compute_5[(cse_var_1 + 79)] = 0f32
+ compute_5[(cse_var_1 + 96)] = 0f32
+ compute_5[(cse_var_1 + 97)] = 0f32
+ compute_5[(cse_var_1 + 98)] = 0f32
+ compute_5[(cse_var_1 + 99)] = 0f32
+ compute_5[(cse_var_1 + 100)] = 0f32
+ compute_5[(cse_var_1 + 101)] = 0f32
+ compute_5[(cse_var_1 + 102)] = 0f32
+ compute_5[(cse_var_1 + 103)] = 0f32
+ compute_5[(cse_var_1 + 104)] = 0f32
+ compute_5[(cse_var_1 + 105)] = 0f32
+ compute_5[(cse_var_1 + 106)] = 0f32
+ compute_5[(cse_var_1 + 107)] = 0f32
+ compute_5[(cse_var_1 + 108)] = 0f32
+ compute_5[(cse_var_1 + 109)] = 0f32
+ compute_5[(cse_var_1 + 110)] = 0f32
+ compute_5[(cse_var_1 + 111)] = 0f32
+ compute_5[(cse_var_1 + 128)] = 0f32
+ compute_5[(cse_var_1 + 129)] = 0f32
+ compute_5[(cse_var_1 + 130)] = 0f32
+ compute_5[(cse_var_1 + 131)] = 0f32
+ compute_5[(cse_var_1 + 132)] = 0f32
+ compute_5[(cse_var_1 + 133)] = 0f32
+ compute_5[(cse_var_1 + 134)] = 0f32
+ compute_5[(cse_var_1 + 135)] = 0f32
+ compute_5[(cse_var_1 + 136)] = 0f32
+ compute_5[(cse_var_1 + 137)] = 0f32
+ compute_5[(cse_var_1 + 138)] = 0f32
+ compute_5[(cse_var_1 + 139)] = 0f32
+ compute_5[(cse_var_1 + 140)] = 0f32
+ compute_5[(cse_var_1 + 141)] = 0f32
+ compute_5[(cse_var_1 + 142)] = 0f32
+ compute_5[(cse_var_1 + 143)] = 0f32
+ compute_5[(cse_var_1 + 160)] = 0f32
+ compute_5[(cse_var_1 + 161)] = 0f32
+ compute_5[(cse_var_1 + 162)] = 0f32
+ compute_5[(cse_var_1 + 163)] = 0f32
+ compute_5[(cse_var_1 + 164)] = 0f32
+ compute_5[(cse_var_1 + 165)] = 0f32
+ compute_5[(cse_var_1 + 166)] = 0f32
+ compute_5[(cse_var_1 + 167)] = 0f32
+ compute_5[(cse_var_1 + 168)] = 0f32
+ compute_5[(cse_var_1 + 169)] = 0f32
+ compute_5[(cse_var_1 + 170)] = 0f32
+ compute_5[(cse_var_1 + 171)] = 0f32
+ compute_5[(cse_var_1 + 172)] = 0f32
+ compute_5[(cse_var_1 + 173)] = 0f32
+ compute_5[(cse_var_1 + 174)] = 0f32
+ compute_5[(cse_var_1 + 175)] = 0f32
+ compute_5[(cse_var_1 + 192)] = 0f32
+ compute_5[(cse_var_1 + 193)] = 0f32
+ compute_5[(cse_var_1 + 194)] = 0f32
+ compute_5[(cse_var_1 + 195)] = 0f32
+ compute_5[(cse_var_1 + 196)] = 0f32
+ compute_5[(cse_var_1 + 197)] = 0f32
+ compute_5[(cse_var_1 + 198)] = 0f32
+ compute_5[(cse_var_1 + 199)] = 0f32
+ compute_5[(cse_var_1 + 200)] = 0f32
+ compute_5[(cse_var_1 + 201)] = 0f32
+ compute_5[(cse_var_1 + 202)] = 0f32
+ compute_5[(cse_var_1 + 203)] = 0f32
+ compute_5[(cse_var_1 + 204)] = 0f32
+ compute_5[(cse_var_1 + 205)] = 0f32
+ compute_5[(cse_var_1 + 206)] = 0f32
+ compute_5[(cse_var_1 + 207)] = 0f32
+ compute_5[(cse_var_1 + 224)] = 0f32
+ compute_5[(cse_var_1 + 225)] = 0f32
+ compute_5[(cse_var_1 + 226)] = 0f32
+ compute_5[(cse_var_1 + 227)] = 0f32
+ compute_5[(cse_var_1 + 228)] = 0f32
+ compute_5[(cse_var_1 + 229)] = 0f32
+ compute_5[(cse_var_1 + 230)] = 0f32
+ compute_5[(cse_var_1 + 231)] = 0f32
+ compute_5[(cse_var_1 + 232)] = 0f32
+ compute_5[(cse_var_1 + 233)] = 0f32
+ compute_5[(cse_var_1 + 234)] = 0f32
+ compute_5[(cse_var_1 + 235)] = 0f32
+ compute_5[(cse_var_1 + 236)] = 0f32
+ compute_5[(cse_var_1 + 237)] = 0f32
+ compute_5[(cse_var_1 + 238)] = 0f32
+ compute_5[(cse_var_1 + 239)] = 0f32
+ for (elem_idx: int32, 0, (placeholder_3[(cse_var_2 + 1)] - placeholder_3[cse_var_2])) {
+ let cse_var_131: int32 = (i.outer.inner*2048)
+ let cse_var_130: int32 = (elem_idx*16)
+ let cse_var_129: int32 = (cse_var_1 + 99)
+ let cse_var_128: int32 = (cse_var_1 + 98)
+ let cse_var_127: int32 = (cse_var_1 + 97)
+ let cse_var_126: int32 = (cse_var_1 + 96)
+ let cse_var_125: int32 = (cse_var_1 + 9)
+ let cse_var_124: int32 = (cse_var_1 + 8)
+ let cse_var_123: int32 = (cse_var_1 + 79)
+ let cse_var_122: int32 = (cse_var_1 + 78)
+ let cse_var_121: int32 = (cse_var_1 + 77)
+ let cse_var_120: int32 = (cse_var_1 + 76)
+ let cse_var_119: int32 = (cse_var_1 + 75)
+ let cse_var_118: int32 = (cse_var_1 + 74)
+ let cse_var_117: int32 = (cse_var_1 + 73)
+ let cse_var_116: int32 = (cse_var_1 + 72)
+ let cse_var_115: int32 = (cse_var_1 + 71)
+ let cse_var_114: int32 = (cse_var_1 + 70)
+ let cse_var_113: int32 = (cse_var_1 + 7)
+ let cse_var_112: int32 = (cse_var_1 + 69)
+ let cse_var_111: int32 = (cse_var_1 + 68)
+ let cse_var_110: int32 = (cse_var_1 + 67)
+ let cse_var_109: int32 = (cse_var_1 + 66)
+ let cse_var_108: int32 = (cse_var_1 + 65)
+ let cse_var_107: int32 = (cse_var_1 + 64)
+ let cse_var_106: int32 = (cse_var_1 + 6)
+ let cse_var_105: int32 = (cse_var_1 + 5)
+ let cse_var_104: int32 = (cse_var_1 + 47)
+ let cse_var_103: int32 = (cse_var_1 + 46)
+ let cse_var_102: int32 = (cse_var_1 + 45)
+ let cse_var_101: int32 = (cse_var_1 + 44)
+ let cse_var_100: int32 = (cse_var_1 + 43)
+ let cse_var_99: int32 = (cse_var_1 + 42)
+ let cse_var_98: int32 = (cse_var_1 + 41)
+ let cse_var_97: int32 = (cse_var_1 + 40)
+ let cse_var_96: int32 = (cse_var_1 + 4)
+ let cse_var_95: int32 = (cse_var_1 + 39)
+ let cse_var_94: int32 = (cse_var_1 + 38)
+ let cse_var_93: int32 = (cse_var_1 + 37)
+ let cse_var_92: int32 = (cse_var_1 + 36)
+ let cse_var_91: int32 = (cse_var_1 + 35)
+ let cse_var_90: int32 = (cse_var_1 + 34)
+ let cse_var_89: int32 = (cse_var_1 + 33)
+ let cse_var_88: int32 = (cse_var_1 + 32)
+ let cse_var_87: int32 = (cse_var_1 + 3)
+ let cse_var_86: int32 = (cse_var_1 + 239)
+ let cse_var_85: int32 = (cse_var_1 + 238)
+ let cse_var_84: int32 = (cse_var_1 + 237)
+ let cse_var_83: int32 = (cse_var_1 + 236)
+ let cse_var_82: int32 = (cse_var_1 + 235)
+ let cse_var_81: int32 = (cse_var_1 + 234)
+ let cse_var_80: int32 = (cse_var_1 + 233)
+ let cse_var_79: int32 = (cse_var_1 + 232)
+ let cse_var_78: int32 = (cse_var_1 + 231)
+ let cse_var_77: int32 = (cse_var_1 + 230)
+ let cse_var_76: int32 = (cse_var_1 + 229)
+ let cse_var_75: int32 = (cse_var_1 + 228)
+ let cse_var_74: int32 = (cse_var_1 + 227)
+ let cse_var_73: int32 = (cse_var_1 + 226)
+ let cse_var_72: int32 = (cse_var_1 + 225)
+ let cse_var_71: int32 = (cse_var_1 + 224)
+ let cse_var_70: int32 = (cse_var_1 + 207)
+ let cse_var_69: int32 = (cse_var_1 + 206)
+ let cse_var_68: int32 = (cse_var_1 + 205)
+ let cse_var_67: int32 = (cse_var_1 + 204)
+ let cse_var_66: int32 = (cse_var_1 + 203)
+ let cse_var_65: int32 = (cse_var_1 + 202)
+ let cse_var_64: int32 = (cse_var_1 + 201)
+ let cse_var_63: int32 = (cse_var_1 + 200)
+ let cse_var_62: int32 = (cse_var_1 + 2)
+ let cse_var_61: int32 = (cse_var_1 + 199)
+ let cse_var_60: int32 = (cse_var_1 + 198)
+ let cse_var_59: int32 = (cse_var_1 + 197)
+ let cse_var_58: int32 = (cse_var_1 + 196)
+ let cse_var_57: int32 = (cse_var_1 + 195)
+ let cse_var_56: int32 = (cse_var_1 + 194)
+ let cse_var_55: int32 = (cse_var_1 + 193)
+ let cse_var_54: int32 = (cse_var_1 + 192)
+ let cse_var_53: int32 = (cse_var_1 + 175)
+ let cse_var_52: int32 = (cse_var_1 + 174)
+ let cse_var_51: int32 = (cse_var_1 + 173)
+ let cse_var_50: int32 = (cse_var_1 + 172)
+ let cse_var_49: int32 = (cse_var_1 + 171)
+ let cse_var_48: int32 = (cse_var_1 + 170)
+ let cse_var_47: int32 = (cse_var_1 + 169)
+ let cse_var_46: int32 = (cse_var_1 + 168)
+ let cse_var_45: int32 = (cse_var_1 + 167)
+ let cse_var_44: int32 = (cse_var_1 + 166)
+ let cse_var_43: int32 = (cse_var_1 + 165)
+ let cse_var_42: int32 = (cse_var_1 + 164)
+ let cse_var_41: int32 = (cse_var_1 + 163)
+ let cse_var_40: int32 = (cse_var_1 + 162)
+ let cse_var_39: int32 = (cse_var_1 + 161)
+ let cse_var_38: int32 = (cse_var_1 + 160)
+ let cse_var_37: int32 = (cse_var_1 + 15)
+ let cse_var_36: int32 = (cse_var_1 + 143)
+ let cse_var_35: int32 = (cse_var_1 + 142)
+ let cse_var_34: int32 = (cse_var_1 + 141)
+ let cse_var_33: int32 = (cse_var_1 + 140)
+ let cse_var_32: int32 = (cse_var_1 + 14)
+ let cse_var_31: int32 = (cse_var_1 + 139)
+ let cse_var_30: int32 = (cse_var_1 + 138)
+ let cse_var_29: int32 = (cse_var_1 + 137)
+ let cse_var_28: int32 = (cse_var_1 + 136)
+ let cse_var_27: int32 = (cse_var_1 + 135)
+ let cse_var_26: int32 = (cse_var_1 + 134)
+ let cse_var_25: int32 = (cse_var_1 + 133)
+ let cse_var_24: int32 = (cse_var_1 + 132)
+ let cse_var_23: int32 = (cse_var_1 + 131)
+ let cse_var_22: int32 = (cse_var_1 + 130)
+ let cse_var_21: int32 = (cse_var_1 + 13)
+ let cse_var_20: int32 = (cse_var_1 + 129)
+ let cse_var_19: int32 = (cse_var_1 + 128)
+ let cse_var_18: int32 = (cse_var_1 + 12)
+ let cse_var_17: int32 = (cse_var_1 + 111)
+ let cse_var_16: int32 = (cse_var_1 + 110)
+ let cse_var_15: int32 = (cse_var_1 + 11)
+ let cse_var_14: int32 = (cse_var_1 + 109)
+ let cse_var_13: int32 = (cse_var_1 + 108)
+ let cse_var_12: int32 = (cse_var_1 + 107)
+ let cse_var_11: int32 = (cse_var_1 + 106)
+ let cse_var_10: int32 = (cse_var_1 + 105)
+ let cse_var_9: int32 = (cse_var_1 + 104)
+ let cse_var_8: int32 = (cse_var_1 + 103)
+ let cse_var_7: int32 = (cse_var_1 + 102)
+ let cse_var_6: int32 = (cse_var_1 + 101)
+ let cse_var_5: int32 = (cse_var_1 + 100)
+ let cse_var_4: int32 = (cse_var_1 + 10)
+ let cse_var_3: int32 = (cse_var_1 + 1)
+ {
+ compute_5[cse_var_1] = (compute_5[cse_var_1] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_3] = (compute_5[cse_var_3] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_62] = (compute_5[cse_var_62] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_87] = (compute_5[cse_var_87] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_96] = (compute_5[cse_var_96] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_105] = (compute_5[cse_var_105] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_106] = (compute_5[cse_var_106] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_113] = (compute_5[cse_var_113] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_124] = (compute_5[cse_var_124] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_125] = (compute_5[cse_var_125] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_4] = (compute_5[cse_var_4] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_21] = (compute_5[cse_var_21] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_32] = (compute_5[cse_var_32] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_37] = (compute_5[cse_var_37] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_88] = (compute_5[cse_var_88] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_89] = (compute_5[cse_var_89] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_90] = (compute_5[cse_var_90] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_91] = (compute_5[cse_var_91] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_92] = (compute_5[cse_var_92] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_93] = (compute_5[cse_var_93] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_94] = (compute_5[cse_var_94] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_95] = (compute_5[cse_var_95] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_97] = (compute_5[cse_var_97] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_98] = (compute_5[cse_var_98] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_99] = (compute_5[cse_var_99] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_100] = (compute_5[cse_var_100] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_101] = (compute_5[cse_var_101] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_102] = (compute_5[cse_var_102] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_103] = (compute_5[cse_var_103] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_104] = (compute_5[cse_var_104] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_107] = (compute_5[cse_var_107] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_108] = (compute_5[cse_var_108] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_109] = (compute_5[cse_var_109] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_110] = (compute_5[cse_var_110] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_111] = (compute_5[cse_var_111] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_112] = (compute_5[cse_var_112] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_114] = (compute_5[cse_var_114] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_115] = (compute_5[cse_var_115] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_116] = (compute_5[cse_var_116] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_117] = (compute_5[cse_var_117] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_118] = (compute_5[cse_var_118] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_119] = (compute_5[cse_var_119] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_120] = (compute_5[cse_var_120] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_121] = (compute_5[cse_var_121] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_122] = (compute_5[cse_var_122] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_123] = (compute_5[cse_var_123] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_126] = (compute_5[cse_var_126] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_127] = (compute_5[cse_var_127] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_128] = (compute_5[cse_var_128] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_129] = (compute_5[cse_var_129] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_19] = (compute_5[cse_var_19] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_20] = (compute_5[cse_var_20] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_22] = (compute_5[cse_var_22] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_23] = (compute_5[cse_var_23] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_24] = (compute_5[cse_var_24] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_25] = (compute_5[cse_var_25] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_26] = (compute_5[cse_var_26] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_27] = (compute_5[cse_var_27] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_28] = (compute_5[cse_var_28] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_29] = (compute_5[cse_var_29] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_30] = (compute_5[cse_var_30] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_31] = (compute_5[cse_var_31] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_33] = (compute_5[cse_var_33] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_34] = (compute_5[cse_var_34] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_35] = (compute_5[cse_var_35] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_36] = (compute_5[cse_var_36] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_38] = (compute_5[cse_var_38] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_39] = (compute_5[cse_var_39] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_40] = (compute_5[cse_var_40] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_41] = (compute_5[cse_var_41] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_42] = (compute_5[cse_var_42] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_43] = (compute_5[cse_var_43] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_44] = (compute_5[cse_var_44] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_45] = (compute_5[cse_var_45] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_46] = (compute_5[cse_var_46] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_47] = (compute_5[cse_var_47] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_48] = (compute_5[cse_var_48] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_49] = (compute_5[cse_var_49] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_50] = (compute_5[cse_var_50] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_51] = (compute_5[cse_var_51] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_52] = (compute_5[cse_var_52] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_53] = (compute_5[cse_var_53] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_54] = (compute_5[cse_var_54] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_55] = (compute_5[cse_var_55] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_56] = (compute_5[cse_var_56] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_57] = (compute_5[cse_var_57] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_58] = (compute_5[cse_var_58] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_59] = (compute_5[cse_var_59] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_60] = (compute_5[cse_var_60] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_61] = (compute_5[cse_var_61] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_63] = (compute_5[cse_var_63] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_64] = (compute_5[cse_var_64] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_65] = (compute_5[cse_var_65] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_66] = (compute_5[cse_var_66] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_67] = (compute_5[cse_var_67] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_68] = (compute_5[cse_var_68] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_69] = (compute_5[cse_var_69] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_70] = (compute_5[cse_var_70] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_71] = (compute_5[cse_var_71] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_72] = (compute_5[cse_var_72] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_73] = (compute_5[cse_var_73] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_74] = (compute_5[cse_var_74] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_75] = (compute_5[cse_var_75] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_76] = (compute_5[cse_var_76] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_77] = (compute_5[cse_var_77] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_78] = (compute_5[cse_var_78] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_79] = (compute_5[cse_var_79] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_80] = (compute_5[cse_var_80] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_81] = (compute_5[cse_var_81] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_82] = (compute_5[cse_var_82] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_83] = (compute_5[cse_var_83] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_84] = (compute_5[cse_var_84] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_85] = (compute_5[cse_var_85] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_86] = (compute_5[cse_var_86] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ }
}
}
}
}
- for (i0.inner: int32, 0, 16) {
- let cse_var_260: int32 = (((i0.outer*8192) + (i0.inner*512)) + (i1.outer*32))
- compute[ramp(cse_var_260, 1, 32)] = max((compute_5[ramp((i0.inner*32), 1, 32)] + placeholder_4[ramp(cse_var_260, 1, 32)]), broadcast(0f32, 32))
+ for (i0.inner: int32, 0, 128) {
+ let cse_var_132: int32 = ((i0.inner*512) + (i0.outer.i1.outer.fused*32))
+ compute[ramp(cse_var_132, 1, 32)] = max((compute_5[ramp((i0.inner*32), 1, 32)] + placeholder_4[ramp(cse_var_132, 1, 32)]), broadcast(0f32, 32))
}
}
}
@@ -1238,7 +855,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 3.392 ms
+ Execution time of this operator: 2.763 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 eb0d9ea40..0427ecd7b 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
=================
-**00:45.665** total execution time for **how_to_tune_with_autotvm** files:
+**00:46.504** 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``) | 00:45.630 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``) | 00:46.468 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``) | 00:00.020 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``) | 00:00.021 | 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 7adcf9ea2..53552978d 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
@@ -1156,8 +1156,8 @@ for this template
TimeoutError
[('tile_f', [-1, 2, 1, 64]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4909501
- No: 9 GFLOPS: 80.77/80.77 result: MeasureResult(costs=(0.0028661723142857144,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6695890426635742, timestamp=1663152848.0716567) [('tile_f', [-1, 1, 4, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5072689
- No: 10 GFLOPS: 0.00/80.77 result: Traceback (most recent call last):
+ No: 9 GFLOPS: 193.79/193.79 result: MeasureResult(costs=(0.0011946226666666668,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.089510440826416, timestamp=1663169826.564597) [('tile_f', [-1, 1, 4, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5072689
+ No: 10 GFLOPS: 0.00/193.79 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
@@ -1280,8 +1280,8 @@ for this template
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, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5092711
- No: 11 GFLOPS: 260.49/260.49 result: MeasureResult(costs=(0.0008887261396648044,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7274281978607178, timestamp=1663152849.0026462) [('tile_f', [-1, 8, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4264713
- No: 12 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+ No: 11 GFLOPS: 260.43/260.43 result: MeasureResult(costs=(0.0008889091049723756,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7412033081054688, timestamp=1663169827.445183) [('tile_f', [-1, 8, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4264713
+ No: 12 GFLOPS: 0.00/260.43 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
@@ -1404,7 +1404,7 @@ for this template
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, 1, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,183542
- No: 13 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+ No: 13 GFLOPS: 0.00/260.43 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
@@ -1527,7 +1527,7 @@ for this template
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, 8, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2482196
- No: 14 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+ No: 14 GFLOPS: 0.00/260.43 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
@@ -1650,9 +1650,9 @@ for this template
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, 64, 1, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10306226
- No: 15 GFLOPS: 5.25/260.49 result: MeasureResult(costs=(0.0440660225,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.857121229171753, timestamp=1663152853.5778544) [('tile_f', [-1, 2, 2, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5330964
- No: 16 GFLOPS: 3.34/260.49 result: MeasureResult(costs=(0.06935714725,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.554947137832642, timestamp=1663152854.8194664) [('tile_f', [-1, 8, 4, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2140058
- No: 17 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+ No: 15 GFLOPS: 5.29/260.43 result: MeasureResult(costs=(0.04376472025000001,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8177464008331299, timestamp=1663169832.0521896) [('tile_f', [-1, 2, 2, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5330964
+ No: 16 GFLOPS: 3.34/260.43 result: MeasureResult(costs=(0.06941178825,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.619394063949585, timestamp=1663169833.2994785) [('tile_f', [-1, 8, 4, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2140058
+ No: 17 GFLOPS: 0.00/260.43 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 142, in build
res = future.result()
File "/usr/lib/python3.7/concurrent/futures/_base.py", line 435, in result
@@ -1670,8 +1670,8 @@ for this template
TimeoutError
[('tile_f', [-1, 2, 2, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10195251
- No: 18 GFLOPS: 28.16/260.49 result: MeasureResult(costs=(0.008221106714285715,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2657346725463867, timestamp=1663152865.8246107) [('tile_f', [-1, 4, 8, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6068603
- No: 19 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+ No: 18 GFLOPS: 28.23/260.43 result: MeasureResult(costs=(0.00819916792857143,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2819502353668213, timestamp=1663169844.3120673) [('tile_f', [-1, 4, 8, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6068603
+ No: 19 GFLOPS: 0.00/260.43 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
@@ -1794,7 +1794,7 @@ for this template
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, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6956993
- No: 20 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+ No: 20 GFLOPS: 0.00/260.43 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
@@ -1973,7 +1973,7 @@ and measure running time.
Best config:
[('tile_f', [-1, 8, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4264713
Finish loading 20 records
- Time cost of this operator: 0.001238
+ Time cost of this operator: 0.001224
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 ec39ce704..9d6ba0947 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
@@ -329,10 +329,10 @@ Timing the untuned program
########## Build without Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs Measurements(us)
--------- --- -------- ------- ----- ------ ------- ----------------
- tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 310.5 98.701 (1, 2, 10, 10, 3) 2 1 [310.5]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.118 0.991 (1, 6, 10, 10) 1 1 [3.118]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.969 0.308 (1, 1, 10, 10, 3) 1 1 [0.969]
- Total_time - 314.587 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 313.3 98.707 (1, 2, 10, 10, 3) 2 1 [313.3]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.121 0.983 (1, 6, 10, 10) 1 1 [3.121]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.984 0.31 (1, 1, 10, 10, 3) 1 1 [0.984]
+ Total_time - 317.405 - - - - -
@@ -398,10 +398,10 @@ 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 245.2 98.819 (1, 1, 10, 10, 6) 2 1 [245.2]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.955 0.788 (1, 6, 10, 10) 1 1 [1.955]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.974 0.392 (1, 1, 10, 10, 3) 1 1 [0.974]
- Total_time - 248.129 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 88.188 96.991 (1, 6, 10, 10, 1) 2 1 [88.188]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.781 1.959 (1, 6, 10, 10) 1 1 [1.781]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.955 1.051 (1, 1, 10, 10, 3) 1 1 [0.955]
+ Total_time - 90.924 - - - - -
diff --git a/docs/_sources/how_to/work_with_microtvm/micro_tflite.rst.txt b/docs/_sources/how_to/work_with_microtvm/micro_tflite.rst.txt
index e26a56ae1..6c044201a 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_tflite.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_tflite.rst.txt
@@ -409,7 +409,7 @@ Now, compile the model for the target:
extern "C"
#endif
TVM_DLL int32_t tvmgen_default_fused_nn_dense_add(void* args, int32_t* arg_type_ids, int32_t num_args, void* out_ret_value, int32_t* out_ret_tcode, void* resource_handle) {
- void* arg_placeholder = (((TVMValue*)args)[0].v_handle);
+ void* arg_p0 = (((TVMValue*)args)[0].v_handle);
- .
- ./codegen
- ./codegen/host
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 da7166722..8df05a81a 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/tmpy02oqjc1/images/random'
+ '/tmp/tmpqbeyr2dj/images/random'
@@ -325,8 +325,8 @@ objects to other stuff? We can display some examples from our datasets using ``m
.. code-block:: none
- /tmp/tmpy02oqjc1/images/target contains 8144 images
- /tmp/tmpy02oqjc1/images/random contains 5000 images
+ /tmp/tmpqbeyr2dj/images/target contains 8144 images
+ /tmp/tmpqbeyr2dj/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: 0.2103 - accuracy: 0.9256 - val_loss: 0.1451 - val_accuracy: 0.9592 - 47s/epoch - 142ms/step
+ 328/328 - 47s - loss: 0.2159 - accuracy: 0.9242 - val_loss: 0.1586 - val_accuracy: 0.9479 - 47s/epoch - 144ms/step
Epoch 2/3
- 328/328 - 43s - loss: 0.0929 - accuracy: 0.9642 - val_loss: 0.1379 - val_accuracy: 0.9577 - 43s/epoch - 132ms/step
+ 328/328 - 44s - loss: 0.0987 - accuracy: 0.9630 - val_loss: 0.1163 - val_accuracy: 0.9592 - 44s/epoch - 133ms/step
Epoch 3/3
- 328/328 - 43s - loss: 0.0636 - accuracy: 0.9761 - val_loss: 0.1846 - val_accuracy: 0.9460 - 43s/epoch - 132ms/step
+ 328/328 - 43s - loss: 0.0668 - accuracy: 0.9752 - val_loss: 0.1528 - val_accuracy: 0.9535 - 43s/epoch - 133ms/step
- <keras.callbacks.History object at 0x7f232bc3b2d0>
+ <keras.callbacks.History object at 0x7ff287c2ab50>
@@ -871,7 +871,7 @@ 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 25.782 seconds)
+ **Total running time of the script:** ( 4 minutes 29.667 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 09471f65a..384676b69 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
@@ -5,16 +5,16 @@
Computation times
=================
-**05:19.716** total execution time for **how_to_work_with_microtvm** files:
+**05:25.473** total execution time for **how_to_work_with_microtvm** files:
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:25.782 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:29.667 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 00:42.520 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 00:44.074 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:08.139 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:08.282 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:03.274 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:03.448 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_microtvm_micro_ethosu.py` (``micro_ethosu.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 f3751d1a8..b30c8c392 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
=================
-**00:42.593** total execution time for **how_to_work_with_relay** files:
+**00:44.630** 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``) | 00:31.230 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:32.950 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:09.902 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:10.172 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.454 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.501 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``) | 00:00.006 | 0.0 MB |
+| :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 888256672..3e1298e8e 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 0x7f235defa9e0>
+ <function my_cuda_math_rule at 0x7ff20fb4d3b0>
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 3216c6e00..1beaedb0c 100644
--- a/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
@@ -5,18 +5,18 @@
Computation times
=================
-**00:07.049** total execution time for **how_to_work_with_schedules** files:
+**00:04.860** total execution time for **how_to_work_with_schedules** files:
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:04.899 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:02.419 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``) | 00:00.937 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``) | 00:01.095 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.526 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.587 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.504 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.574 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.099 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.101 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.042 | 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 dafb9ec80..ae3f32045 100644
--- a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
@@ -347,7 +347,7 @@ The importing needs to happen before the tensorized GEMV being executed.
C: Buffer(C_2: Pointer(float32), float32, [524288], [])}
buffer_map = {A_1: A, B_1: B, C_1: C}
preflattened_buffer_map = {A_1: A_3: Buffer(A_2, float32, [1024, 64], []), B_1: B_3: Buffer(B_2, float32, [512, 64], []), C_1: C_3: Buffer(C_2, float32, [1024, 512], [])} {
- attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmpycd4q_qg/input0.cc'\nsource_filename = \"/tmp/tmpycd4q_qg/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/tmpzj0yygjo/input0.cc'\nsource_filename = \"/tmp/tmpzj0yygjo/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 cd3736681..66991e96e 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
=================
-**00:21.766** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:22.227** total execution time for **topic_vta_tutorials_autotvm** files:
+---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:21.760 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:22.221 | 0.0 MB |
+---------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``) | 00:00.006 | 0.0 MB |
+---------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
index e57802cbb..1a453b577 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
@@ -291,7 +291,7 @@ 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 23.35s!
+ resnet18_v1 inference graph built in 23.84s!
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 40ac0022e..379acfceb 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
@@ -335,7 +335,7 @@ The compilation steps are:
"target_host parameter is going to be deprecated. "
/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 16.52s!
+ yolov3-tiny inference graph built in 16.70s!
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 35ec1e4b6..57f92993d 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
=================
-**01:32.997** total execution time for **topic_vta_tutorials_frontend** files:
+**01:34.330** total execution time for **topic_vta_tutorials_frontend** files:
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:49.267 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:49.948 | 0.0 MB |
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:43.730 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:44.382 | 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 6fe0d70d3..717ce38da 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
=================
-**00:02.965** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.144** total execution time for **topic_vta_tutorials_optimize** files:
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.584 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.711 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.380 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.433 | 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 4aaa4e134..481e32b4a 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
=================
-**00:00.700** total execution time for **topic_vta_tutorials** files:
+**00:00.820** total execution time for **topic_vta_tutorials** files:
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.376 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.436 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.324 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.384 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
index 8e057ca85..314c308e8 100644
--- a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
@@ -207,8 +207,8 @@ trials, we can load the best schedule from the log file and apply it.
.. code-block:: none
- *E
+ *E
@@ -333,7 +333,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 93.187 ms
+ Execution time of this operator: 94.199 ms
@@ -451,7 +451,7 @@ operations.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 8.233 seconds)
+ **Total running time of the script:** ( 1 minutes 9.315 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 1e2e0faa0..a3b703353 100644
--- a/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
@@ -462,16 +462,16 @@ reduce variance, we take 5 measurements and average them.
waiting for device...
device available
Get devices for measurement successfully!
- No: 1 GFLOPS: 10.80/10.80 result: MeasureResult(costs=(0.0248441218,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5323035717010498, timestamp=1663151621.0319817) [('tile_y', [-1, 1]), ('tile_x', [-1, 256])],None,80
- No: 2 GFLOPS: 2.94/10.80 result: MeasureResult(costs=(0.09143316,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6016459465026855, timestamp=1663151623.178534) [('tile_y', [-1, 4]), ('tile_x', [-1, 8])],None,32
- No: 3 GFLOPS: 11.81/11.81 result: MeasureResult(costs=(0.0227219188,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5910851955413818, timestamp=1663151623.7432706) [('tile_y', [-1, 64]), ('tile_x', [-1, 32])],None,56
- No: 4 GFLOPS: 1.87/11.81 result: MeasureResult(costs=(0.1435910904,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.419503927230835, timestamp=1663151626.7323549) [('tile_y', [-1, 1]), ('tile_x', [-1, 4])],None,20
- No: 5 GFLOPS: 3.70/11.81 result: MeasureResult(costs=(0.0724911754,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.3038039207458496, timestamp=1663151628.1694636) [('tile_y', [-1, 256]), ('tile_x', [-1, 16])],None,48
- No: 6 GFLOPS: 1.82/11.81 result: MeasureResult(costs=(0.1474662836,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.570136070251465, timestamp=1663151630.78386) [('tile_y', [-1, 512]), ('tile_x', [-1, 4])],None,29
- No: 7 GFLOPS: 0.87/11.81 result: MeasureResult(costs=(0.306912871,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.029702663421631, timestamp=1663151636.3859282) [('tile_y', [-1, 512]), ('tile_x', [-1, 2])],None,19
- No: 8 GFLOPS: 10.69/11.81 result: MeasureResult(costs=(0.0251120858,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5538527965545654, timestamp=1663151636.9501348) [('tile_y', [-1, 4]), ('tile_x', [-1, 64])],None,62
- No: 9 GFLOPS: 1.63/11.81 result: MeasureResult(costs=(0.16428153139999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.7408392429351807, timestamp=1663151639.811811) [('tile_y', [-1, 2]), ('tile_x', [-1, 2])],None,11
- No: 10 GFLOPS: 2.68/11.81 result: MeasureResult(costs=(0.1002093654,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7228963375091553, timestamp=1663151641.5799263) [('tile_y', [-1, 4]), ('tile_x', [-1, 4])],None,22
+ No: 1 GFLOPS: 10.58/10.58 result: MeasureResult(costs=(0.0253738562,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5411367416381836, timestamp=1663168558.0770411) [('tile_y', [-1, 1]), ('tile_x', [-1, 256])],None,80
+ No: 2 GFLOPS: 2.96/10.58 result: MeasureResult(costs=(0.0907982816,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6053133010864258, timestamp=1663168560.2534258) [('tile_y', [-1, 4]), ('tile_x', [-1, 8])],None,32
+ No: 3 GFLOPS: 11.72/11.72 result: MeasureResult(costs=(0.022896776,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6365985870361328, timestamp=1663168560.8492706) [('tile_y', [-1, 64]), ('tile_x', [-1, 32])],None,56
+ No: 4 GFLOPS: 1.83/11.72 result: MeasureResult(costs=(0.1463097326,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.460474729537964, timestamp=1663168563.9087136) [('tile_y', [-1, 1]), ('tile_x', [-1, 4])],None,20
+ No: 5 GFLOPS: 3.63/11.72 result: MeasureResult(costs=(0.0738881412,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.3340740203857422, timestamp=1663168565.3735304) [('tile_y', [-1, 256]), ('tile_x', [-1, 16])],None,48
+ No: 6 GFLOPS: 1.76/11.72 result: MeasureResult(costs=(0.1526215432,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.612694025039673, timestamp=1663168568.0252585) [('tile_y', [-1, 512]), ('tile_x', [-1, 4])],None,29
+ No: 7 GFLOPS: 0.87/11.72 result: MeasureResult(costs=(0.307587342,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.056328296661377, timestamp=1663168573.6724384) [('tile_y', [-1, 512]), ('tile_x', [-1, 2])],None,19
+ No: 8 GFLOPS: 10.54/11.72 result: MeasureResult(costs=(0.0254762176,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5553841590881348, timestamp=1663168574.2455058) [('tile_y', [-1, 4]), ('tile_x', [-1, 64])],None,62
+ No: 9 GFLOPS: 1.62/11.72 result: MeasureResult(costs=(0.16578761,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.75015926361084, timestamp=1663168577.1159546) [('tile_y', [-1, 2]), ('tile_x', [-1, 2])],None,11
+ No: 10 GFLOPS: 2.61/11.72 result: MeasureResult(costs=(0.1029824636,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7545139789581299, timestamp=1663168578.9281523) [('tile_y', [-1, 4]), ('tile_x', [-1, 4])],None,22
diff --git a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
index 42fe7bea2..5114ee0d7 100644
--- a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
@@ -327,7 +327,7 @@ standard deviation.
.. code-block:: none
- {'mean': 510.0764587099957, 'median': 509.4831571499981, 'std': 1.2992827880861189}
+ {'mean': 512.7538775699941, 'median': 512.935857499906, 'std': 1.6370233811732053}
@@ -563,30 +563,30 @@ the tuning data to.
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
-
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 17.53/ 17.53 GFLOPS | Progress: (4/20) | 6.86 s
[Task 1/25] Current/Best: 6.11/ 17.53 GFLOPS | Progress: (8/20) | 9.31 s
[Task 1/25] Current/Best: 11.21/ 22.29 GFLOPS | Progress: (12/20) | 11.82 s
[Task 1/25] Current/Best: 16.48/ 22.29 GFLOPS | Progress: (16/20) | 13.52 s
[Task 1/25] Current/Best: 11.34/ 23.64 GFLOPS | Progress: (20/20) | 15.30 s Done.
-
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 11.92/ 12.34 GFLOPS | Progress: (4/20) | 3.81 s
[Task 2/25] Current/Best: 12.48/ 18.59 GFLOPS | Progress: (8/20) | 5.10 s
[Task 2/25] Current/Best: 20.91/ 20.91 GFLOPS | Progress: (12/20) | 6.41 s
[Task 2/25] Current/Best: 11.09/ 20.91 GFLOPS | Progress: (16/20) | 7.71 s
[Task 2/25] Current/Best: 17.08/ 20.91 GFLOPS | Progress: (20/20) | 9.33 s Done.
-
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 1.63/ 9.93 GFLOPS | Progress: (4/20) | 5.89 s
[Task 3/25] Current/Best: 15.33/ 16.85 GFLOPS | Progress: (8/20) | 7.84 s
[Task 3/25] Current/Best: 15.04/ 16.85 GFLOPS | Progress: (12/20) | 9.61 s
[Task 3/25] Current/Best: 6.85/ 23.01 GFLOPS | Progress: (16/20) | 11.62 s
[Task 3/25] Current/Best: 11.05/ 23.01 GFLOPS | Progress: (20/20) | 16.22 s Done.
-
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 9.02/ 18.75 GFLOPS | Progress: (4/20) | 2.44 s
[Task 4/25] Current/Best: 6.46/ 18.75 GFLOPS | Progress: (8/20) | 6.80 s
[Task 4/25] Current/Best: 20.28/ 20.28 GFLOPS | Progress: (12/20) | 11.22 s
[Task 4/25] Current/Best: 16.50/ 20.39 GFLOPS | Progress: (16/20) | 13.44 s
[Task 4/25] Current/Best: 12.56/ 20.39 GFLOPS | Progress: (20/20) | 15.32 s Done.
-
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 8.91/ 9.93 GFLOPS | Progress: (4/20) | 2.64 s
[Task 5/25] Current/Best: 11.55/ 11.55 GFLOPS | Progress: (8/20) | 4.72 s
[Task 5/25] Current/Best: 11.60/ 17.95 GFLOPS | Progress: (12/20) | 7.81 s
[Task 5/25] Current/Best: 11.64/ 22.26 GFLOPS | Progress: (16/20) | 9.24 s
[Task 5/25] Current/Best: 12.07/ 22.26 GFLOPS | Progress: (20/20) | 11.11 s Done.
-
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 12.02/ 19.92 GFLOPS | Progress: (4/20) | 4.01 s
[Task 6/25] Current/Best: 18.90/ 19.92 GFLOPS | Progress: (8/20) | 5.77 s
[Task 6/25] Current/Best: 13.27/ 19.92 GFLOPS | Progress: (12/20) | 7.75 s
[Task 6/25] Current/Best: 19.35/ 19.92 GFLOPS | Progress: (16/20) | 10.02 s
[Task 6/25] Current/Best: 3.73/ 19.92 GFLOPS | Progress: (20/20) | 12.61 s Done.
-
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 9.79/ 12.13 GFLOPS | Progress: (4/20) | 3.70 s
[Task 7/25] Current/Best: 19.52/ 20.00 GFLOPS | Progress: (8/20) | 5.25 s
[Task 7/25] Current/Best: 16.19/ 20.00 GFLOPS | Progress: (12/20) | 7.17 s
[Task 7/25] Current/Best: 12.19/ 20.22 GFLOPS | Progress: (16/20) | 9.25 s
[Task 7/25] Current/Best: 5.96/ 20.49 GFLOPS | Progress: (20/20) | 11.77 s Done.
-
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 10.00/ 13.59 GFLOPS | Progress: (4/20) | 2.93 s
[Task 8/25] Current/Best: 9.58/ 13.59 GFLOPS | Progress: (8/20) | 7.62 s
[Task 8/25] Current/Best: 12.72/ 13.76 GFLOPS | Progress: (12/20) | 13.77 s
[Task 8/25] Current/Best: 18.99/ 18.99 GFLOPS | Progress: (16/20) | 15.91 s
[Task 8/25] Current/Best: 18.69/ 18.99 GFLOPS | Progress: (20/20) | 22.39 s Done.
-
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 14.32/ 14.32 GFLOPS | Progress: (4/20) | 11.98 s
[Task 9/25] Current/Best: 22.81/ 22.81 GFLOPS | Progress: (8/20) | 13.78 s
[Task 9/25] Current/Best: 7.74/ 22.81 GFLOPS | Progress: (12/20) | 16.15 s
[Task 9/25] Current/Best: 17.59/ 22.81 GFLOPS | Progress: (16/20) | 18.80 s
[Task 9/25] Current/Best: 9.05/ 22.81 GFLOPS | Progress: (20/20) | 26.41 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 18.09/ 18.09 GFLOPS | Progress: (4/20) | 2.60 s
[Task 10/25] Current/Best: 15.62/ 18.09 GFLOPS | Progress: (8/20) | 4.23 s
[Task 10/25] Current/Best: 11.53/ 18.94 GFLOPS | Progress: (12/20) | 5.76 s
[Task 10/25] Current/Best: 19.03/ 19.87 GFLOPS | Progress: (16/20) | 6.87 s
[Task 10/25] Current/Best: 8.25/ 19.87 GFLOPS | Progress: (20/20
) | 8.42 s Done.
-
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 10.64/ 18.18 GFLOPS | Progress: (4/20) | 3.41 s
[Task 11/25] Current/Best: 14.88/ 18.18 GFLOPS | Progress: (8/20) | 6.15 s
[Task 11/25] Current/Best: 15.96/ 18.18 GFLOPS | Progress: (12/20) | 8.26 s
[Task 11/25] Current/Best: 11.90/ 20.68 GFLOPS | Progress: (16/20) | 11.00 s
[Task 11/25] Current/Best: 17.48/ 20.68 GFLOPS | Progress: (20/20) | 13.05 s Done.
-
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 7.77/ 17.87 GFLOPS | Progress: (4/20) | 5.43 s
[Task 12/25] Current/Best: 5.14/ 17.87 GFLOPS | Progress: (8/20) | 9.15 s
[Task 12/25] Current/Best: 18.73/ 18.73 GFLOPS | Progress: (12/20) | 11.16 s
[Task 12/25] Current/Best: 15.06/ 18.73 GFLOPS | Progress: (16/20) | 13.92 s
[Task 12/25] Current/Best: 15.09/ 18.73 GFLOPS | Progress: (20/20) | 15.87 s Done.
-
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 8.40/ 17.30 GFLOPS | Progress: (4/20) | 3.71 s
[Task 13/25] Current/Best: 15.18/ 20.62 GFLOPS | Progress: (8/20) | 6.18 s
[Task 13/25] Current/Best: 18.69/ 21.59 GFLOPS | Progress: (12/20) | 9.11 s
[Task 13/25] Current/Best: 12.19/ 21.59 GFLOPS | Progress: (16/20) | 12.51 s
[Task 13/25] Current/Best: 17.83/ 21.59 GFLOPS | Progress: (20/20) | 14.77 s Done.
-
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 13.05/ 13.37 GFLOPS | Progress: (4/20) | 3.37 s
[Task 14/25] Current/Best: 6.07/ 13.37 GFLOPS | Progress: (8/20) | 5.55 s
[Task 14/25] Current/Best: 19.25/ 19.25 GFLOPS | Progress: (12/20) | 8.10 s
[Task 14/25] Current/Best: 16.24/ 19.25 GFLOPS | Progress: (16/20) | 9.79 s Done.
-
[Task 14/25] Current/Best: 16.87/ 19.25 GFLOPS | Progress: (20/20) | 11.52 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 15.61/ 17.07 GFLOPS | Progress: (4/20) | 2.75 s
[Task 15/25] Current/Best: 12.69/ 17.70 GFLOPS | Progress: (8/20) | 4.13 s
[Task 15/25] Current/Best: 9.97/ 21.71 GFLOPS | Progress: (12/20) | 6.19 s
[Task 15/25] Current/Best: 20.36/ 21.71 GFLOPS | Progress: (16/20) | 9.13 s
[Task 15/25] Current/Best: 9.50/ 21.71 GFLOPS | Progress: (20/20) | 10.11 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 18.10/ 18.10 GFLOPS | Progress: (4/20) | 2.98 s
[Task 16/25] Current/Best: 3.03/ 18.10 GFLOPS | Progress: (8/20) | 4.61 s
[Task 16/25] Current/Best: 17.89/ 19.42 GFLOPS | Progress: (12/20) | 5.83 s
[Task 16/25] Current/Best: 17.90/ 19.42 GFLOPS | Progress: (16/20) |
7.20 s
[Task 16/25] Current/Best: 9.83/ 21.09 GFLOPS | Progress: (20/20) | 9.26 s Done.
-
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 12.61/ 16.13 GFLOPS | Progress: (4/20) | 4.77 s
[Task 17/25] Current/Best: 12.57/ 22.93 GFLOPS | Progress: (8/20) | 7.55 s
[Task 17/25] Current/Best: 16.40/ 22.93 GFLOPS | Progress: (12/20) | 9.67 s
[Task 17/25] Current/Best: 16.43/ 22.93 GFLOPS | Progress: (16/20) | 11.80 s
[Task 17/25] Current/Best: 9.68/ 22.93 GFLOPS | Progress: (20/20) | 13.94 s Done.
-
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 10.60/ 16.45 GFLOPS | Progress: (4/20) | 3.74 s
[Task 18/25] Current/Best: 10.59/ 19.20 GFLOPS | Progress: (8/20) | 7.20 s
[Task 18/25] Current/Best: 18.35/ 19.20 GFLOPS | Progress: (12/20) | 9.16 s
[Task 18/25] Current/Best: 9.93/ 19.20 GFLOPS | Progress: (16/20) | 12.71 s
[Task 18/25] Current/Best: 20.44/ 20.44 GFLOPS | Progress: (20/20) | 14.26 s Done.
-
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 7.15/ 19.77 GFLOPS | Progress: (4/20) | 6.05 s
[Task 19/25] Current/Best: 2.69/ 19.77 GFLOPS | Progress: (8/20) | 9.28 s
[Task 19/25] Current/Best: 18.40/ 20.68 GFLOPS | Progress: (12/20) | 12.06 s
[Task 19/25] Current/Best: 13.22/ 20.68 GFLOPS | Progress: (16/20) | 14.95 s
[Task 19/25] Current/Best: 2.69/ 21.59 GFLOPS | Progress: (20/20) | 17.82 s Done.
-
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 9.29/ 15.48 GFLOPS | Progress: (4/20) | 3.30 s Done.
+
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 17.50/ 17.50 GFLOPS | Progress: (4/20) | 6.43 s
[Task 1/25] Current/Best: 6.08/ 17.50 GFLOPS | Progress: (8/20) | 9.46 s
[Task 1/25] Current/Best: 11.21/ 22.29 GFLOPS | Progress: (12/20) | 11.95 s
[Task 1/25] Current/Best: 16.51/ 22.29 GFLOPS | Progress: (16/20) | 13.65 s
[Task 1/25] Current/Best: 11.30/ 23.49 GFLOPS | Progress: (20/20) | 15.44 s Done.
+
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 12.25/ 12.29 GFLOPS | Progress: (4/20) | 3.71 s
[Task 2/25] Current/Best: 12.52/ 17.94 GFLOPS | Progress: (8/20) | 5.04 s
[Task 2/25] Current/Best: 20.99/ 20.99 GFLOPS | Progress: (12/20) | 6.41 s
[Task 2/25] Current/Best: 10.90/ 20.99 GFLOPS | Progress: (16/20) | 7.71 s
[Task 2/25] Current/Best: 16.80/ 20.99 GFLOPS | Progress: (20/20) | 9.28 s Done.
+
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 1.63/ 10.07 GFLOPS | Progress: (4/20) | 5.92 s
[Task 3/25] Current/Best: 15.35/ 16.84 GFLOPS | Progress: (8/20) | 7.88 s
[Task 3/25] Current/Best: 15.00/ 16.84 GFLOPS | Progress: (12/20) | 9.62 s
[Task 3/25] Current/Best: 6.79/ 22.71 GFLOPS | Progress: (16/20) | 11.62 s
[Task 3/25] Current/Best: 11.01/ 22.71 GFLOPS | Progress: (20/20) | 16.23 s Done.
+
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 8.92/ 18.63 GFLOPS | Progress: (4/20) | 2.49 s
[Task 4/25] Current/Best: 6.56/ 18.63 GFLOPS | Progress: (8/20) | 6.91 s
[Task 4/25] Current/Best: 20.42/ 20.42 GFLOPS | Progress: (12/20) | 11.41 s
[Task 4/25] Current/Best: 16.20/ 20.42 GFLOPS | Progress: (16/20) | 13.68 s
[Task 4/25] Current/Best: 12.96/ 20.42 GFLOPS | Progress: (20/20) | 15.71 s Done.
+
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 9.04/ 9.73 GFLOPS | Progress: (4/20) | 2.69 s
[Task 5/25] Current/Best: 11.61/ 11.61 GFLOPS | Progress: (8/20) | 4.78 s
[Task 5/25] Current/Best: 10.94/ 18.09 GFLOPS | Progress: (12/20) | 7.89 s
[Task 5/25] Current/Best: 11.44/ 21.99 GFLOPS | Progress: (16/20) | 9.31 s
[Task 5/25] Current/Best: 11.98/ 21.99 GFLOPS | Progress: (20/20) | 11.16 s Done.
+
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 11.93/ 19.94 GFLOPS | Progress: (4/20) | 4.04 s
[Task 6/25] Current/Best: 18.85/ 19.94 GFLOPS | Progress: (8/20) | 5.82 s
[Task 6/25] Current/Best: 13.18/ 19.94 GFLOPS | Progress: (12/20) | 7.83 s
[Task 6/25] Current/Best: 19.45/ 19.94 GFLOPS | Progress: (16/20) | 10.11 s
[Task 6/25] Current/Best: 3.71/ 19.94 GFLOPS | Progress: (20/20) | 12.70 s Done.
+
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 9.62/ 12.11 GFLOPS | Progress: (4/20) | 3.69 s
[Task 7/25] Current/Best: 18.76/ 19.89 GFLOPS | Progress: (8/20) | 5.25 s
[Task 7/25] Current/Best: 15.69/ 19.89 GFLOPS | Progress: (12/20) | 7.21 s
[Task 7/25] Current/Best: 12.12/ 19.89 GFLOPS | Progress: (16/20) | 9.32 s
[Task 7/25] Current/Best: 6.05/ 20.40 GFLOPS | Progress: (20/20) | 11.84 s Done.
+
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 10.19/ 13.75 GFLOPS | Progress: (4/20) | 3.00 s
[Task 8/25] Current/Best: 9.57/ 13.75 GFLOPS | Progress: (8/20) | 7.90 s
[Task 8/25] Current/Best: 13.22/ 13.75 GFLOPS | Progress: (12/20) | 14.16 s
[Task 8/25] Current/Best: 19.05/ 19.05 GFLOPS | Progress: (16/20) | 16.27 s
[Task 8/25] Current/Best: 18.67/ 19.05 GFLOPS | Progress: (20/20) | 22.87 s Done.
+
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 14.24/ 14.24 GFLOPS | Progress: (4/20) | 12.03 s
[Task 9/25] Current/Best: 22.38/ 22.38 GFLOPS | Progress: (8/20) | 13.90 s
[Task 9/25] Current/Best: 7.95/ 22.38 GFLOPS | Progress: (12/20) | 16.29 s
[Task 9/25] Current/Best: 17.80/ 22.38 GFLOPS | Progress: (16/20) | 19.00 s
[Task 9/25] Current/Best: 8.88/ 22.38 GFLOPS | Progress: (20/20) | 26.71 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 18.24/ 18.24 GFLOPS | Progress: (4/20) | 2.64 s
[Task 10/25] Current/Best: 15.79/ 18.24 GFLOPS | Progress: (8/20) | 4.24 s
[Task 10/25] Current/Best: 11.35/ 19.01 GFLOPS | Progress: (12/20) | 5.80 s
[Task 10/25] Current/Best: 19.04/ 20.34 GFLOPS | Progress: (16/20) | 6.92 s
[Task 10/25] Current/Best: 8.44/ 20.34 GFLOPS | Progress: (20/20
) | 8.51 s Done.
+
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 10.73/ 18.12 GFLOPS | Progress: (4/20) | 3.46 s
[Task 11/25] Current/Best: 14.84/ 18.12 GFLOPS | Progress: (8/20) | 6.26 s
[Task 11/25] Current/Best: 15.91/ 18.12 GFLOPS | Progress: (12/20) | 8.38 s
[Task 11/25] Current/Best: 11.79/ 20.62 GFLOPS | Progress: (16/20) | 11.17 s
[Task 11/25] Current/Best: 18.65/ 20.62 GFLOPS | Progress: (20/20) | 13.25 s Done.
+
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 7.76/ 17.93 GFLOPS | Progress: (4/20) | 5.42 s
[Task 12/25] Current/Best: 5.03/ 17.93 GFLOPS | Progress: (8/20) | 9.18 s
[Task 12/25] Current/Best: 18.84/ 18.84 GFLOPS | Progress: (12/20) | 11.21 s
[Task 12/25] Current/Best: 14.98/ 18.84 GFLOPS | Progress: (16/20) | 14.06 s
[Task 12/25] Current/Best: 15.16/ 18.84 GFLOPS | Progress: (20/20) | 16.03 s Done.
+
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 8.27/ 17.30 GFLOPS | Progress: (4/20) | 3.78 s
[Task 13/25] Current/Best: 15.08/ 20.55 GFLOPS | Progress: (8/20) | 6.25 s
[Task 13/25] Current/Best: 18.73/ 21.73 GFLOPS | Progress: (12/20) | 9.21 s
[Task 13/25] Current/Best: 12.19/ 21.73 GFLOPS | Progress: (16/20) | 12.69 s
[Task 13/25] Current/Best: 17.44/ 21.73 GFLOPS | Progress: (20/20) | 15.06 s Done.
+
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 11.66/ 13.18 GFLOPS | Progress: (4/20) | 3.44 s
[Task 14/25] Current/Best: 6.03/ 13.18 GFLOPS | Progress: (8/20) | 5.65 s
[Task 14/25] Current/Best: 19.33/ 19.33 GFLOPS | Progress: (12/20) | 8.25 s
[Task 14/25] Current/Best: 15.23/ 19.33 GFLOPS | Progress: (16/20) | 9.97 s Done.
+
[Task 14/25] Current/Best: 16.98/ 19.33 GFLOPS | Progress: (20/20) | 11.77 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 15.23/ 17.11 GFLOPS | Progress: (4/20) | 2.81 s
[Task 15/25] Current/Best: 12.63/ 17.81 GFLOPS | Progress: (8/20) | 4.14 s
[Task 15/25] Current/Best: 9.75/ 20.38 GFLOPS | Progress: (12/20) | 6.25 s
[Task 15/25] Current/Best: 20.33/ 20.38 GFLOPS | Progress: (16/20) | 9.48 s
[Task 15/25] Current/Best: 9.47/ 20.38 GFLOPS | Progress: (20/20) | 10.52 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 18.83/ 18.83 GFLOPS | Progress: (4/20) | 3.18 s
[Task 16/25] Current/Best: 3.03/ 18.83 GFLOPS | Progress: (8/20) | 4.82 s
[Task 16/25] Current/Best: 17.47/ 19.34 GFLOPS | Progress: (12/20) | 6.07 s
[Task 16/25] Current/Best: 17.63/ 19.34 GFLOPS | Progress: (16/20) |
7.45 s
[Task 16/25] Current/Best: 10.17/ 21.02 GFLOPS | Progress: (20/20) | 9.53 s Done.
+
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 12.55/ 16.06 GFLOPS | Progress: (4/20) | 4.86 s
[Task 17/25] Current/Best: 13.06/ 21.99 GFLOPS | Progress: (8/20) | 7.78 s
[Task 17/25] Current/Best: 16.47/ 21.99 GFLOPS | Progress: (12/20) | 9.92 s
[Task 17/25] Current/Best: 16.41/ 21.99 GFLOPS | Progress: (16/20) | 12.08 s
[Task 17/25] Current/Best: 9.95/ 21.99 GFLOPS | Progress: (20/20) | 14.23 s Done.
+
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 10.53/ 16.32 GFLOPS | Progress: (4/20) | 3.82 s
[Task 18/25] Current/Best: 10.58/ 18.41 GFLOPS | Progress: (8/20) | 7.34 s
[Task 18/25] Current/Best: 18.90/ 18.90 GFLOPS | Progress: (12/20) | 9.32 s
[Task 18/25] Current/Best: 9.88/ 18.90 GFLOPS | Progress: (16/20) | 12.93 s
[Task 18/25] Current/Best: 20.46/ 20.46 GFLOPS | Progress: (20/20) | 14.50 s Done.
+
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 6.31/ 19.52 GFLOPS | Progress: (4/20) | 6.27 s
[Task 19/25] Current/Best: 2.69/ 19.52 GFLOPS | Progress: (8/20) | 9.55 s
[Task 19/25] Current/Best: 17.88/ 19.84 GFLOPS | Progress: (12/20) | 12.37 s
[Task 19/25] Current/Best: 13.43/ 20.18 GFLOPS | Progress: (16/20) | 15.23 s
[Task 19/25] Current/Best: 2.69/ 21.94 GFLOPS | Progress: (20/20) | 18.04 s Done.
+
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 8.04/ 14.90 GFLOPS | Progress: (4/20) | 3.46 s Done.
Done.
-
[Task 20/25] Current/Best: 9.98/ 15.48 GFLOPS | Progress: (8/20) | 6.77 s
[Task 20/25] Current/Best: 2.32/ 15.48 GFLOPS | Progress: (12/20) | 10.71 s
[Task 20/25] Current/Best: 10.98/ 15.48 GFLOPS | Progress: (16/20) | 14.31 s
[Task 20/25] Current/Best: 11.76/ 21.72 GFLOPS | Progress: (20/20) | 16.45 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 6.34/ 17.77 GFLOPS | Progress: (4/20) | 3.24 s
[Task 21/25] Current/Best: 14.63/ 17.77 GFLOPS | Progress: (8/20) | 4.84 s
[Task 21/25] Current/Best: 1.61/ 17.77 GFLOPS | Progress: (12/20) | 6.99 s
[Task 21/25] Current/Best: 15.98/ 17.77 GFLOPS | Progress: (16/20) | 10.44 s
[Task 21/25] Current/Best: 4.46/ 17.77 GFLOPS | Progress: (20/20) | 17.57 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 2.70/ 16.94 GFLOPS | Progress: (4/20
) | 2.69 s
[Task 22/25] Current/Best: 8.88/ 21.16 GFLOPS | Progress: (8/20) | 4.66 s
[Task 22/25] Current/Best: 19.82/ 21.16 GFLOPS | Progress: (12/20) | 6.95 s
[Task 22/25] Current/Best: 15.56/ 21.16 GFLOPS | Progress: (16/20) | 9.00 s
[Task 22/25] Current/Best: 13.91/ 21.16 GFLOPS | Progress: (20/20) | 10.68 s Done.
-
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 16.62/ 19.62 GFLOPS | Progress: (4/20) | 3.31 s
[Task 23/25] Current/Best: 13.88/ 19.88 GFLOPS | Progress: (8/20) | 6.69 s
[Task 23/25] Current/Best: 20.66/ 21.70 GFLOPS | Progress: (12/20) | 8.52 s
[Task 23/25] Current/Best: 6.58/ 21.70 GFLOPS | Progress: (16/20) | 15.38 s
[Task 23/25] Current/Best: 7.59/ 21.70 GFLOPS | Progress: (20/20) | 19.60 s Done.
-
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 8.43/ 8.43 GFLOPS | Progress: (4/20) | 11.79 s
[Task 24/25] Current/Best: 1.93/ 8.43 GFLOPS | Progress: (8/20) | 22.81 s
[Task 24/25] Current/Best: 3.72/ 8.43 GFLOPS | Progress: (12/20) | 34.37 s Done.
-
[Task 24/25] Current/Best: 6.56/ 8.99 GFLOPS | Progress: (16/20) | 39.70 s
[Task 24/25] Current/Best: 2.95/ 8.99 GFLOPS | Progress: (20/20) | 45.60 s Done.
-
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 25/25] Current/Best: 1.55/ 2.80 GFLOPS | Progress: (4/20) | 11.61 s
[Task 25/25] Current/Best: 5.93/ 7.82 GFLOPS | Progress: (8/20) | 22.89 s
[Task 25/25] Current/Best: 5.96/ 7.82 GFLOPS | Progress: (12/20) | 34.36 s
[Task 25/25] Current/Best: 5.76/ 8.75 GFLOPS | Progress: (16/20) | 36.10 s
[Task 25/25] Current/Best: 2.91/ 8.94 GFLOPS | Progress: (20/20) | 46.78 s
+
[Task 20/25] Current/Best: 9.43/ 14.90 GFLOPS | Progress: (8/20) | 7.00 s
[Task 20/25] Current/Best: 2.34/ 14.90 GFLOPS | Progress: (12/20) | 10.98 s
[Task 20/25] Current/Best: 11.02/ 14.90 GFLOPS | Progress: (16/20) | 14.64 s
[Task 20/25] Current/Best: 11.89/ 21.26 GFLOPS | Progress: (20/20) | 16.76 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 6.33/ 17.57 GFLOPS | Progress: (4/20) | 3.29 s
[Task 21/25] Current/Best: 14.54/ 17.57 GFLOPS | Progress: (8/20) | 4.88 s
[Task 21/25] Current/Best: 1.61/ 17.57 GFLOPS | Progress: (12/20) | 7.08 s
[Task 21/25] Current/Best: 16.07/ 17.57 GFLOPS | Progress: (16/20) | 10.62 s
[Task 21/25] Current/Best: 4.44/ 17.57 GFLOPS | Progress: (20/20) | 17.90 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 2.70/ 16.95 GFLOPS | Progress: (4/20
) | 2.76 s
[Task 22/25] Current/Best: 9.32/ 21.05 GFLOPS | Progress: (8/20) | 4.74 s
[Task 22/25] Current/Best: 19.76/ 21.05 GFLOPS | Progress: (12/20) | 7.09 s
[Task 22/25] Current/Best: 13.78/ 21.05 GFLOPS | Progress: (16/20) | 9.16 s
[Task 22/25] Current/Best: 13.22/ 21.05 GFLOPS | Progress: (20/20) | 10.94 s Done.
+
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 16.47/ 16.47 GFLOPS | Progress: (4/20) | 3.40 s
[Task 23/25] Current/Best: 14.21/ 19.74 GFLOPS | Progress: (8/20) | 6.78 s
[Task 23/25] Current/Best: 20.40/ 21.37 GFLOPS | Progress: (12/20) | 8.61 s
[Task 23/25] Current/Best: 6.32/ 21.37 GFLOPS | Progress: (16/20) | 15.77 s
[Task 23/25] Current/Best: 7.18/ 21.37 GFLOPS | Progress: (20/20) | 20.07 s Done.
+
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 8.61/ 8.61 GFLOPS | Progress: (4/20) | 11.91 s
[Task 24/25] Current/Best: 1.89/ 8.61 GFLOPS | Progress: (8/20) | 23.01 s
[Task 24/25] Current/Best: 3.70/ 8.61 GFLOPS | Progress: (12/20) | 34.61 s Done.
+
[Task 24/25] Current/Best: 6.53/ 8.85 GFLOPS | Progress: (16/20) | 40.11 s
[Task 24/25] Current/Best: 2.91/ 8.85 GFLOPS | Progress: (20/20) | 46.17 s Done.
+
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 25/25] Current/Best: 1.54/ 2.83 GFLOPS | Progress: (4/20) | 11.67 s
[Task 25/25] Current/Best: 5.51/ 7.53 GFLOPS | Progress: (8/20) | 23.02 s
[Task 25/25] Current/Best: 5.78/ 7.53 GFLOPS | Progress: (12/20) | 34.55 s
[Task 25/25] Current/Best: 5.80/ 9.05 GFLOPS | Progress: (16/20) | 36.42 s
[Task 25/25] Current/Best: 2.81/ 9.05 GFLOPS | Progress: (20/20) | 47.15 s
@@ -690,8 +690,8 @@ Verify that the optimized model runs and produces the same results:
.. code-block:: none
- class='n02123045 tabby, tabby cat' with probability=0.621105
- class='n02123159 tiger cat' with probability=0.356377
+ class='n02123045 tabby, tabby cat' with probability=0.621104
+ class='n02123159 tiger cat' with probability=0.356378
class='n02124075 Egyptian cat' with probability=0.019712
class='n02129604 tiger, Panthera tigris' with probability=0.001215
class='n04040759 radiator' with probability=0.000262
@@ -748,8 +748,8 @@ improvement in comparing the optimized model to the unoptimized model.
.. code-block:: none
- optimized: {'mean': 408.13420215000406, 'median': 407.84730635000415, 'std': 0.8589507690660168}
- unoptimized: {'mean': 510.0764587099957, 'median': 509.4831571499981, 'std': 1.2992827880861189}
+ optimized: {'mean': 410.2872488299545, 'median': 410.1678336997793, 'std': 1.1521201837318131}
+ unoptimized: {'mean': 512.7538775699941, 'median': 512.935857499906, 'std': 1.6370233811732053}
@@ -772,7 +772,7 @@ profiling/benchmarking.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 10 minutes 16.261 seconds)
+ **Total running time of the script:** ( 10 minutes 27.786 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 c68d550d9..f2ac1cde9 100644
--- a/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
+++ b/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
@@ -282,7 +282,7 @@ device and returns the measured cost. Network overhead is excluded.
.. code-block:: none
- 1.218e-07 secs/op
+ 1.259e-07 secs/op
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index ab28a052d..462192e4c 100644
--- a/docs/_sources/tutorial/intro_topi.rst.txt
+++ b/docs/_sources/tutorial/intro_topi.rst.txt
@@ -263,7 +263,7 @@ As you can see, scheduled stages of computation have been accumulated and we can
.. code-block:: none
- [stage(a, placeholder(a, 0x203f4fc0)), stage(b, placeholder(b, 0x4ba2db0)), 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, 0x218b92c0)), stage(b, placeholder(b, 0x55a8650)), 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 [...]
diff --git a/docs/_sources/tutorial/sg_execution_times.rst.txt b/docs/_sources/tutorial/sg_execution_times.rst.txt
index 987c18b51..856f25e36 100644
--- a/docs/_sources/tutorial/sg_execution_times.rst.txt
+++ b/docs/_sources/tutorial/sg_execution_times.rst.txt
@@ -5,32 +5,32 @@
Computation times
=================
-**13:19.653** total execution time for **tutorial** files:
+**13:36.736** total execution time for **tutorial** files:
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 10:16.261 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 10:27.786 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:08.233 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:09.315 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 00:58.719 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 01:01.778 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:31.088 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:31.768 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:23.986 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:24.418 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.704 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:00.779 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:00.511 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.719 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.142 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.163 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``) | 00:00.004 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``) | 00:00.005 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_uma.py` (``uma.py``) | 00:00.002 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``) | 00:00.001 | 0.0 MB |
-+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_tvmc_python.py` (``tvmc_python.py``) | 00:00.001 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
+| :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``) | 00:00.001 | 0.0 MB |
++------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_install.py` (``install.py``) | 00:00.001 | 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 d2002ec1b..ee5b741f5 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -512,10 +512,10 @@ We can now compare the different schedules
.. code-block:: none
Operator Timing Performance
- numpy 6.994579998718109e-06 1.0
- naive 6.686100000000001e-06 0.9558972806409187
- parallel 6.903500000000001e-06 0.986978489239554
- vector 2.4511599999999998e-05 3.504370527535923
+ numpy 7.2787500175763855e-06 1.0
+ naive 6.686000000000001e-06 0.9185643117094227
+ parallel 7.0276e-06 0.9654954467497964
+ vector 2.4614e-05 3.3816245839688492
@@ -936,7 +936,7 @@ matrix multiplication.
.. code-block:: none
- Numpy running time: 0.017814
+ Numpy running time: 0.019387
@@ -996,7 +996,7 @@ optimizations.
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- none: 3.254258
+ none: 3.429760
@@ -1101,7 +1101,7 @@ schedule.
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- blocking: 0.293253
+ blocking: 0.324998
@@ -1199,7 +1199,7 @@ already cache friendly from our previous optimizations.
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- vectorization: 0.338814
+ vectorization: 0.346217
@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, [1048576], []),
@@ -1275,7 +1275,7 @@ more cache friendly.
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- loop permutation: 0.117879
+ loop permutation: 0.124102
@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, [1048576], []),
@@ -1376,7 +1376,7 @@ optimized schedule.
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- array packing: 0.110081
+ array packing: 0.109818
@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, [1048576], []),
@@ -1471,7 +1471,7 @@ to `C` when all the block results are ready.
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- block caching: 0.110496
+ block caching: 0.110771
@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, [1048576], []),
@@ -1559,7 +1559,7 @@ of thread-level parallelization.
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- parallelization: 0.145947
+ parallelization: 0.147587
@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, [1048576], []),
@@ -1640,13 +1640,13 @@ working, we can compare the results.
.. code-block:: none
Operator Timing Performance
- none 3.2542575046000004 1.0
- blocking 0.2932525351 0.09011350044840577
- vectorization 0.3388140416 0.10411408474009054
- loop permutation 0.1178787782 0.036222941188081906
- array packing 0.11008126830000001 0.033826846260443896
- block caching 0.11049627980000001 0.03395437504371116
- parallelization 0.1459466946 0.04484792441707503
+ none 3.4297597029999998 1.0
+ blocking 0.3249975128 0.09475809996709848
+ vectorization 0.34621678740000006 0.10094491083359727
+ loop permutation 0.12410205590000001 0.03618389235591296
+ array packing 0.1098184735 0.03201929085700731
+ block caching 0.11077079940000001 0.03229695634452441
+ parallelization 0.1475874085 0.04303141365003086
@@ -1686,6 +1686,11 @@ operations with tunable parameters that allows you to automatically optimize
the computation for specific platforms.
+.. rst-class:: sphx-glr-timing
+
+ **Total running time of the script:** ( 1 minutes 1.778 seconds)
+
+
.. _sphx_glr_download_tutorial_tensor_expr_get_started.py:
.. only:: html
diff --git a/docs/commit_hash b/docs/commit_hash
index 1f1dc0775..ff2415b09 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-2aa0d1fbfcf4a31e343cc6852fdc4abd660c850a
+a40849342d250bd585e19434e4a2473fcf978bcb
diff --git a/docs/how_to/compile_models/from_darknet.html b/docs/how_to/compile_models/from_darknet.html
index bc84a6c2f..8adb69d37 100644
--- a/docs/how_to/compile_models/from_darknet.html
+++ b/docs/how_to/compile_models/from_darknet.html
@@ -574,7 +574,7 @@ class:['truck 0.9266'] left:471 top:83 right:689 bottom:169
class:['bicycle 0.9984'] left:111 top:113 right:577 bottom:447
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 3.162 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 5.675 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 e223b199d..7b99ab5c9 100644
--- a/docs/how_to/compile_models/from_keras.html
+++ b/docs/how_to/compile_models/from_keras.html
@@ -493,7 +493,7 @@ 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 945ms/step
+1/1 [==============================] - 1s 963ms/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 de6e5bbe7..926b4da16 100644
--- a/docs/how_to/compile_models/from_mxnet.html
+++ b/docs/how_to/compile_models/from_mxnet.html
@@ -427,7 +427,7 @@ to download the full example code</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"x"</span><span class="p">,</span> <a href="https://docs.python.org/3/library/stdtypes.html#tuple" title="builtins.tuple" class="sphx-glr-backref-module-builtins sphx-glr-backref-type-py-class sphx-glr-backref-instance"><span class="n">x</span><span class="o">.</span><span class="n">shape</span></a><span class="p">)</span>
</pre></div>
</div>
-<img src="../../_images/sphx_glr_from_mxnet_001.png" srcset="../../_images/sphx_glr_from_mxnet_001.png" alt="from mxnet" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip616bf519-d84d-4695-a579-d4c9f206c94b 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.zip12d54117-6adf-40be-b795-3e3081d46c5e 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 f1a12e863..47898a27d 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -435,13 +435,12 @@ Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdo
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading: "https://oneflow-public.oss-cn-beijing.aliyuncs.com/model_zoo/flowvision/classification/ResNet/resnet18.zip" to /workspace/.oneflow/flowvision_cache/resnet18.zip
0%| | 0.00/41.5M [00:00<?, ?B/s]
- 15%|#5 | 6.33M/41.5M [00:00<00:00, 60.1MB/s]
- 29%|##9 | 12.1M/41.5M [00:00<00:00, 39.0MB/s]
- 45%|####5 | 18.8M/41.5M [00:00<00:00, 49.8MB/s]
- 58%|#####7 | 24.1M/41.5M [00:00<00:00, 48.0MB/s]
- 77%|#######7 | 32.0M/41.5M [00:00<00:00, 53.2MB/s]
- 96%|#########6| 40.0M/41.5M [00:00<00:00, 57.9MB/s]
-100%|##########| 41.5M/41.5M [00:00<00:00, 54.8MB/s]
+ 19%|#9 | 7.99M/41.5M [00:00<00:00, 67.0MB/s]
+ 39%|###8 | 16.0M/41.5M [00:00<00:00, 70.6MB/s]
+ 58%|#####7 | 24.0M/41.5M [00:00<00:00, 47.4MB/s]
+ 77%|#######7 | 32.0M/41.5M [00:00<00:00, 50.3MB/s]
+ 90%|########9 | 37.2M/41.5M [00:00<00:00, 46.8MB/s]
+100%|##########| 41.5M/41.5M [00:00<00:00, 47.8MB/s]
</pre></div>
</div>
</div>
diff --git a/docs/how_to/compile_models/from_pytorch.html b/docs/how_to/compile_models/from_pytorch.html
index 8ea244780..47ad2b732 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -414,8 +414,9 @@ be unstable.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>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]
- 44%|####3 | 19.6M/44.7M [00:00<00:00, 205MB/s]
-100%|##########| 44.7M/44.7M [00:00<00:00, 242MB/s]
+ 39%|###9 | 17.6M/44.7M [00:00<00:00, 185MB/s]
+ 93%|#########3| 41.6M/44.7M [00:00<00:00, 224MB/s]
+100%|##########| 44.7M/44.7M [00:00<00:00, 220MB/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 4a61c7446..e26bd4a39 100644
--- a/docs/how_to/compile_models/from_tensorflow.html
+++ b/docs/how_to/compile_models/from_tensorflow.html
@@ -636,7 +636,7 @@ 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 6.729 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 9.617 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 8e23b443c..88b616679 100644
--- a/docs/how_to/compile_models/sg_execution_times.html
+++ b/docs/how_to/compile_models/sg_execution_times.html
@@ -327,7 +327,7 @@
<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>05:08.365</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>05:21.075</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 81%" />
@@ -336,43 +336,43 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="from_tensorflow.html#sphx-glr-how-to-compile-models-from-tensorflow-py"><span class="std std-ref">Compile Tensorflow Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tensorflow.py</span></code>)</p></td>
-<td><p>01:06.729</p></td>
+<td><p>01:09.617</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="from_darknet.html#sphx-glr-how-to-compile-models-from-darknet-py"><span class="std std-ref">Compile YOLO-V2 and YOLO-V3 in DarkNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_darknet.py</span></code>)</p></td>
-<td><p>01:03.162</p></td>
+<td><p>01:05.675</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>00:39.177</p></td>
+<td><p>00:41.350</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>00:28.318</p></td>
+<td><p>00:28.858</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>00:25.761</p></td>
+<td><p>00:27.428</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><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>00:24.882</p></td>
+<td><p>00:24.557</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="from_coreml.html#sphx-glr-how-to-compile-models-from-coreml-py"><span class="std std-ref">Compile CoreML Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_coreml.py</span></code>)</p></td>
-<td><p>00:21.841</p></td>
+<td><p>00:23.199</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>00:19.904</p></td>
+<td><p>00:20.335</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>00:16.112</p></td>
+<td><p>00:17.210</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>00:02.480</p></td>
+<td><p>00:02.846</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 73f508873..e47536c61 100644
--- a/docs/how_to/deploy_models/deploy_model_on_android.html
+++ b/docs/how_to/deploy_models/deploy_model_on_android.html
@@ -653,7 +653,7 @@ to the remote android device.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 15.9743 15.9829 16.0650 15.8152 0.0614
+ 16.1038 16.0971 16.1992 16.0414 0.0496
</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 efee439d2..943259330 100644
--- a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
+++ b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
@@ -436,14 +436,42 @@ be unstable.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>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]
- 9%|9 | 16.0M/170M [00:00<00:00, 168MB/s]
- 22%|##2 | 37.9M/170M [00:00<00:00, 204MB/s]
- 35%|###5 | 60.0M/170M [00:00<00:00, 217MB/s]
- 49%|####8 | 82.5M/170M [00:00<00:00, 224MB/s]
- 61%|######1 | 104M/170M [00:00<00:00, 215MB/s]
- 73%|#######3 | 125M/170M [00:00<00:00, 188MB/s]
- 88%|########8 | 150M/170M [00:00<00:00, 211MB/s]
-100%|##########| 170M/170M [00:00<00:00, 213MB/s]
+ 2%|2 | 3.57M/170M [00:00<00:04, 37.4MB/s]
+ 4%|4 | 7.35M/170M [00:00<00:04, 38.7MB/s]
+ 7%|6 | 11.0M/170M [00:00<00:04, 35.8MB/s]
+ 10%|9 | 16.9M/170M [00:00<00:03, 45.3MB/s]
+ 13%|#2 | 21.9M/170M [00:00<00:03, 47.6MB/s]
+ 16%|#5 | 26.5M/170M [00:00<00:04, 35.8MB/s]
+ 18%|#7 | 30.3M/170M [00:00<00:04, 36.3MB/s]
+ 21%|## | 35.2M/170M [00:00<00:03, 40.4MB/s]
+ 24%|##3 | 40.0M/170M [00:01<00:03, 42.9MB/s]
+ 26%|##6 | 44.3M/170M [00:01<00:03, 41.0MB/s]
+ 30%|##9 | 50.1M/170M [00:01<00:02, 46.6MB/s]
+ 32%|###2 | 55.1M/170M [00:01<00:02, 48.3MB/s]
+ 35%|###5 | 59.9M/170M [00:01<00:02, 47.5MB/s]
+ 38%|###8 | 64.8M/170M [00:01<00:02, 48.6MB/s]
+ 41%|####1 | 70.3M/170M [00:01<00:02, 51.3MB/s]
+ 44%|####4 | 75.2M/170M [00:01<00:02, 48.8MB/s]
+ 47%|####7 | 79.9M/170M [00:01<00:01, 47.3MB/s]
+ 50%|####9 | 84.5M/170M [00:02<00:02, 41.5MB/s]
+ 52%|#####2 | 88.8M/170M [00:02<00:02, 42.3MB/s]
+ 55%|#####4 | 93.0M/170M [00:02<00:01, 42.4MB/s]
+ 57%|#####7 | 97.1M/170M [00:02<00:01, 39.3MB/s]
+ 59%|#####9 | 101M/170M [00:02<00:01, 37.3MB/s]
+ 62%|######1 | 105M/170M [00:02<00:01, 38.5MB/s]
+ 64%|######4 | 109M/170M [00:02<00:01, 40.3MB/s]
+ 67%|######7 | 114M/170M [00:02<00:01, 43.6MB/s]
+ 71%|####### | 120M/170M [00:02<00:01, 47.6MB/s]
+ 73%|#######3 | 124M/170M [00:03<00:01, 43.9MB/s]
+ 76%|#######5 | 129M/170M [00:03<00:00, 43.2MB/s]
+ 78%|#######8 | 133M/170M [00:03<00:00, 40.4MB/s]
+ 81%|########1 | 138M/170M [00:03<00:00, 45.1MB/s]
+ 84%|########4 | 143M/170M [00:03<00:00, 46.6MB/s]
+ 88%|########7 | 149M/170M [00:03<00:00, 49.9MB/s]
+ 91%|#########1| 155M/170M [00:03<00:00, 53.0MB/s]
+ 94%|#########4| 160M/170M [00:03<00:00, 54.9MB/s]
+ 97%|#########7| 166M/170M [00:03<00:00, 51.1MB/s]
+100%|##########| 170M/170M [00:04<00:00, 44.2MB/s]
/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:3878: 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)
/usr/local/lib/python3.7/dist-packages/torchvision/models/detection/anchor_utils.py:127: 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').
@@ -541,7 +569,7 @@ torchvision rcnn models.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Get 9 valid boxes
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 57.889 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 8.175 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 5d0b4a2de..09663b7e5 100644
--- a/docs/how_to/deploy_models/deploy_prequantized.html
+++ b/docs/how_to/deploy_models/deploy_prequantized.html
@@ -480,7 +480,8 @@ training. Other models require a full post training calibration.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>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]
-100%|##########| 13.6M/13.6M [00:00<00:00, 188MB/s]
+ 63%|######3 | 8.59M/13.6M [00:00<00:00, 90.0MB/s]
+100%|##########| 13.6M/13.6M [00:00<00:00, 108MB/s]
</pre></div>
</div>
</div>
@@ -569,7 +570,7 @@ 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)
- 90.1986 90.0857 94.0330 89.9495 0.4338
+ 90.4280 90.3556 91.4413 90.1687 0.2261
</pre></div>
</div>
<div class="admonition note">
@@ -608,7 +609,7 @@ 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 10.116 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 11.984 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 67c76f162..dcd7864ba 100644
--- a/docs/how_to/deploy_models/deploy_prequantized_tflite.html
+++ b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
@@ -573,7 +573,7 @@ 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)
- 120.0377 119.9317 122.4276 119.3749 0.4780
+ 122.5429 122.5292 124.1027 121.9173 0.3705
</pre></div>
</div>
<div class="admonition note">
@@ -601,7 +601,7 @@ 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> ( 1 minutes 54.857 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 55.655 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 51cb1b9e7..7033d29f7 100644
--- a/docs/how_to/deploy_models/deploy_quantized.html
+++ b/docs/how_to/deploy_models/deploy_quantized.html
@@ -509,7 +509,7 @@ 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 20.366 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 24.558 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 1ecca1e1f..f4a6e546d 100644
--- a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
+++ b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
@@ -441,25 +441,23 @@ to your device.</p>
Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/ssd_512_resnet50_v1_voc-9c8b225a.zip...
0%| | 0/132723 [00:00<?, ?KB/s]
- 4%|4 | 5372/132723 [00:00<00:02, 53716.47KB/s]
- 10%|9 | 12665/132723 [00:00<00:01, 65008.43KB/s]
- 15%|#5 | 19948/132723 [00:00<00:01, 68577.37KB/s]
- 21%|## | 27267/132723 [00:00<00:01, 70396.61KB/s]
- 26%|##6 | 34604/132723 [00:00<00:01, 71465.66KB/s]
- 32%|###1 | 42043/132723 [00:00<00:01, 72448.79KB/s]
- 37%|###7 | 49437/132723 [00:00<00:01, 72932.56KB/s]
- 43%|####2 | 56833/132723 [00:00<00:01, 73255.51KB/s]
- 48%|####8 | 64174/132723 [00:00<00:00, 73300.65KB/s]
- 54%|#####3 | 71541/132723 [00:01<00:00, 73413.50KB/s]
- 59%|#####9 | 78883/132723 [00:01<00:00, 73381.35KB/s]
- 65%|######4 | 86244/132723 [00:01<00:00, 73450.00KB/s]
- 71%|####### | 93591/132723 [00:01<00:00, 73453.37KB/s]
- 76%|#######6 | 100964/132723 [00:01<00:00, 73528.83KB/s]
- 82%|########1 | 108344/132723 [00:01<00:00, 73604.01KB/s]
- 87%|########7 | 115705/132723 [00:01<00:00, 73480.31KB/s]
- 93%|#########2| 123091/132723 [00:01<00:00, 73592.93KB/s]
- 98%|#########8| 130451/132723 [00:01<00:00, 73492.71KB/s]
-100%|##########| 132723/132723 [00:01<00:00, 72445.75KB/s]
+ 5%|4 | 6061/132723 [00:00<00:02, 60605.00KB/s]
+ 11%|# | 14279/132723 [00:00<00:01, 73291.78KB/s]
+ 16%|#6 | 21609/132723 [00:00<00:01, 65455.35KB/s]
+ 23%|##2 | 29950/132723 [00:00<00:01, 72063.22KB/s]
+ 29%|##8 | 38148/132723 [00:00<00:01, 75478.19KB/s]
+ 35%|###5 | 46525/132723 [00:00<00:01, 78211.55KB/s]
+ 41%|#### | 54408/132723 [00:00<00:01, 78091.50KB/s]
+ 47%|####7 | 62829/132723 [00:00<00:00, 80004.30KB/s]
+ 53%|#####3 | 70861/132723 [00:00<00:00, 79510.75KB/s]
+ 60%|#####9 | 79182/132723 [00:01<00:00, 80636.24KB/s]
+ 66%|######6 | 87601/132723 [00:01<00:00, 81712.50KB/s]
+ 72%|#######2 | 96039/132723 [00:01<00:00, 82516.09KB/s]
+ 79%|#######8 | 104432/132723 [00:01<00:00, 82939.84KB/s]
+ 85%|########5 | 112850/132723 [00:01<00:00, 83312.31KB/s]
+ 91%|#########1| 121186/132723 [00:01<00:00, 72318.24KB/s]
+ 98%|#########7| 129463/132723 [00:01<00:00, 75154.44KB/s]
+100%|##########| 132723/132723 [00:01<00:00, 77096.56KB/s]
</pre></div>
</div>
<p>Create TVM runtime and do inference
@@ -502,7 +500,7 @@ 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> ( 2 minutes 38.412 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> ( 2 minutes 42.368 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 fd2426829..e6322b9c2 100644
--- a/docs/how_to/deploy_models/sg_execution_times.html
+++ b/docs/how_to/deploy_models/sg_execution_times.html
@@ -327,7 +327,7 @@
<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>11:16.656</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>11:40.421</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 86%" />
@@ -336,35 +336,35 @@
</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>02:57.889</p></td>
+<td><p>03:08.175</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>02:38.412</p></td>
+<td><p>02:42.368</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>01:54.857</p></td>
+<td><p>01:55.655</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>01:20.366</p></td>
+<td><p>01:24.558</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>01:10.116</p></td>
+<td><p>01:11.984</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>00:30.446</p></td>
+<td><p>00:31.951</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>00:22.595</p></td>
+<td><p>00:23.137</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>00:21.968</p></td>
+<td><p>00:22.586</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 96633df90..02544472b 100644
--- a/docs/how_to/extend_tvm/bring_your_own_datatypes.html
+++ b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
@@ -612,7 +612,7 @@ 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.zip2311a753-afdc-4829-87a6-73829e254cc7 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.zip670083a9-8dde-4f6e-bbc7-64ce068df60e 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 2786a2fd5..acbb311ae 100644
--- a/docs/how_to/extend_tvm/sg_execution_times.html
+++ b/docs/how_to/extend_tvm/sg_execution_times.html
@@ -327,7 +327,7 @@
<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>00:41.558</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:41.775</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -336,15 +336,15 @@
</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>00:38.390</p></td>
+<td><p>00:38.553</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>00:02.213</p></td>
+<td><p>00:02.245</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>00:00.947</p></td>
+<td><p>00:00.969</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 bdb6356ba..00c69673a 100644
--- a/docs/how_to/extend_tvm/use_pass_instrument.html
+++ b/docs/how_to/extend_tvm/use_pass_instrument.html
@@ -512,10 +512,10 @@ 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: 6729us [6729us] (45.91%; 45.91%)
-FoldScaleAxis: 7929us [5us] (54.09%; 54.09%)
- FoldConstant: 7923us [1676us] (54.06%; 99.93%)
- InferType: 6247us [6247us] (42.62%; 78.84%)
+InferType: 6757us [6757us] (45.94%; 45.94%)
+FoldScaleAxis: 7953us [6us] (54.06%; 54.06%)
+ FoldConstant: 7947us [1671us] (54.02%; 99.93%)
+ InferType: 6276us [6276us] (42.67%; 78.98%)
</pre></div>
</div>
</div>
@@ -537,10 +537,10 @@ 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: 6313us [6313us] (42.74%; 42.74%)
-FoldScaleAxis: 8459us [5us] (57.26%; 57.26%)
- FoldConstant: 8453us [1653us] (57.23%; 99.94%)
- InferType: 6800us [6800us] (46.03%; 80.44%)
+InferType: 6362us [6362us] (44.65%; 44.65%)
+FoldScaleAxis: 7885us [5us] (55.35%; 55.35%)
+ FoldConstant: 7880us [1682us] (55.31%; 99.94%)
+ InferType: 6198us [6198us] (43.51%; 78.66%)
</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 8d02478fd..6379a3f1c 100644
--- a/docs/how_to/optimize_operators/opt_conv_cuda.html
+++ b/docs/how_to/optimize_operators/opt_conv_cuda.html
@@ -564,7 +564,7 @@ latency of convolution.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Convolution: </span><span class="si">%f</span><span class="s2"> ms"</span> <span class="o">%</span> <span class="p">(</span><span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">w</span><span class="p">,</span> <span class="n">b</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span> <span class="o">*</span> <span cl [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 54.210207 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 37.319583 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 e0f9fbab3..5c8dd55cc 100644
--- a/docs/how_to/optimize_operators/opt_conv_tensorcore.html
+++ b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
@@ -906,7 +906,7 @@ be able to run on our build server</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"conv2d with tensor core: </span><span class="si">%f</span><span class="s2"> ms"</span> <span class="o">%</span> <span class="p">(</span><span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">w</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span> <span class="o">* [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 7.739008 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 12.897894 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 b758c246c..b4a637bea 100644
--- a/docs/how_to/optimize_operators/opt_gemm.html
+++ b/docs/how_to/optimize_operators/opt_gemm.html
@@ -461,8 +461,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
<span class="nb">print</span><span class="p">(</span><span class="s2">"Baseline: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018910
-Baseline: 3.256800
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.019168
+Baseline: 3.438547
</pre></div>
</div>
<p>In TVM, we can always inspect lower level IR to debug or optimize our schedule.
@@ -522,7 +522,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt1: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.310534
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.328304
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -589,7 +589,7 @@ vastly.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt2: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.351062
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.350647
</pre></div>
</div>
<p>Here is the generated IR after vectorization.</p>
@@ -650,7 +650,7 @@ the access pattern for A matrix is more cache friendly.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt3: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.113828
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.121355
</pre></div>
</div>
<p>Here is the generated IR after loop permutation.</p>
@@ -733,7 +733,7 @@ flattening.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt4: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.108118
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109599
</pre></div>
</div>
<p>Here is the generated IR after array packing.</p>
@@ -819,7 +819,7 @@ write to C when all the block results are ready.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt5: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.111650
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.112299
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -909,7 +909,7 @@ write to C when all the block results are ready.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt6: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">opt6_time</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.146985
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.148837
</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 5cfc524d4..524e85f37 100644
--- a/docs/how_to/optimize_operators/sg_execution_times.html
+++ b/docs/how_to/optimize_operators/sg_execution_times.html
@@ -327,7 +327,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-optimize-operators-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:34.361</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:35.649</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -336,15 +336,15 @@
</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>00:32.054</p></td>
+<td><p>00:33.063</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>00:01.244</p></td>
+<td><p>00:01.410</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>00:01.062</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 ba2f6ce58..2b78977d8 100644
--- a/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
+++ b/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
@@ -327,7 +327,7 @@
<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>06:23.136</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>06:35.777</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 85%" />
@@ -336,27 +336,27 @@
</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>03:17.760</p></td>
+<td><p>03:24.880</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>01:23.325</p></td>
+<td><p>01:24.368</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>00:47.730</p></td>
+<td><p>00:56.966</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>00:36.545</p></td>
+<td><p>00:31.401</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><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>00:08.980</p></td>
+<td><p>00:09.149</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><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>00:08.797</p></td>
+<td><p>00:09.012</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 bdeac7b32..2e17eabe7 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
@@ -1004,7 +1004,7 @@ 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: 0.367 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.353 ms
</pre></div>
</div>
</div>
@@ -1567,7 +1567,7 @@ In the example below we resume the status and do more 5 trials.</p>
Get devices for measurement successfully!
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 17.760 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 24.880 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_arm.html b/docs/how_to/tune_with_autoscheduler/tune_network_arm.html
index 8a579f56e..1fd3b498c 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_arm.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_arm.html
@@ -604,196 +604,196 @@ The task scheduler will just optimize this objective.</p>
Extract tasks...
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
-========== Task 0 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 1024], [1, 1, 1024, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 1024]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 1024, 1024]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 1024]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 0 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 1024], [1, 1, 1024, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 1024]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 1024, 1024]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 1024]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 1 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 256], [1, 1, 256, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 1 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 256], [1, 1, 256, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 2 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 256]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 256, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 2 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 256]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 256, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 3 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 128], [1, 1, 128, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 128, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 3 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 128], [1, 1, 128, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 128, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 4 (workload key: ["d7b65649a4dd54becea0a52aabbc5af5", [1, 1000], [1, 1000]]) ==========
-placeholder = PLACEHOLDER [1, 1000]
-T_softmax_maxelem(i0) max= placeholder[i0, k]
-T_softmax_exp(i0, i1) = tir.exp((placeholder[i0, i1] - T_softmax_maxelem[i0]))
+========== Task 4 (workload key: ["7d79c516e212fe1d73f5dbb90eaca2cf", [1, 1000], [1, 1000]]) ==========
+p0 = PLACEHOLDER [1, 1000]
+T_softmax_maxelem(i0) max= p0[i0, k]
+T_softmax_exp(i0, i1) = tir.exp((p0[i0, i1] - T_softmax_maxelem[i0]))
T_softmax_expsum(i0) += T_softmax_exp[i0, k]
T_softmax_norm(i0, i1) = (T_softmax_exp[i0, i1]/T_softmax_expsum[i0])
-========== Task 5 (workload key: ["69115f188984ae34ede37c3b8ca40b43", [1, 7, 7, 1024], [1, 1, 1, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 1024]
-tensor(ax0, ax1, ax2, ax3) += placeholder[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
+========== Task 5 (workload key: ["be3babb9a46e32f66b717a3e2a2d522c", [1, 7, 7, 1024], [1, 1, 1, 1024]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 1024]
+tensor(ax0, ax1, ax2, ax3) += p0[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
tensor(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3]/(float32((select((bool)1, ((ax1 + 1)*7), (((ax1 + 1)*7) + 1)) - (ax1*7)))*float32((select((bool)1, ((ax2 + 1)*7), (((ax2 + 1)*7) + 1)) - (ax2*7)))))
-========== Task 6 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 512], [1, 1, 512, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 1024]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 1024]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 6 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 512], [1, 1, 512, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 1024]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 1024]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 7 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 128], [1, 1, 128, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 128]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 128, 128]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 7 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 128], [1, 1, 128, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 128]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 128, 128]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 8 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 128]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 128, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 8 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 128]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 128, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 9 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 256]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 256, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 9 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 256]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 256, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 10 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 512]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 512, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 10 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 512]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 512, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 11 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 112, 112, 64], [3, 3, 64, 1], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 112, 112, 64]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 64, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 11 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 112, 112, 64], [3, 3, 64, 1], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 112, 112, 64]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 64, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 12 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 64], [1, 1, 64, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 64, 128]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 12 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 64], [1, 1, 64, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 64, 128]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 13 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 128]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 128, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 13 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 128]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 128, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 14 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 512], [1, 1, 512, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 14 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 512], [1, 1, 512, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 15 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 112, 112, 32], [3, 3, 32, 1], [1, 1, 1, 32], [1, 112, 112, 32]]) ==========
-placeholder = PLACEHOLDER [1, 112, 112, 32]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 32, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 32]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 15 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 112, 112, 32], [3, 3, 32, 1], [1, 1, 1, 32], [1, 112, 112, 32]]) ==========
+p0 = PLACEHOLDER [1, 112, 112, 32]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 32, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 32]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 16 (workload key: ["2ca148ecea6508ce625f85719021344f", [1, 224, 224, 3], [3, 3, 3, 32], [1, 112, 1, 1], [1, 112, 1, 1], [1, 112, 112, 32]]) ==========
-placeholder = PLACEHOLDER [1, 224, 224, 3]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 225)) && (i2 >= 1)) && (i2 < 225)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 3, 32]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 112, 1, 1]
-T_multiply(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3]*placeholder[ax0, ax1, 0, 0])
-placeholder = PLACEHOLDER [1, 112, 1, 1]
-T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, 0, 0])
+========== Task 16 (workload key: ["ad24d4d2f83975ff580a4833fbf3ef94", [1, 224, 224, 3], [3, 3, 3, 32], [1, 112, 1, 1], [1, 112, 1, 1], [1, 112, 112, 32]]) ==========
+p0 = PLACEHOLDER [1, 224, 224, 3]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 225)) && (i2 >= 1)) && (i2 < 225)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 3, 32]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 112, 1, 1]
+T_multiply(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3]*p2[ax0, ax1, 0, 0])
+p3 = PLACEHOLDER [1, 112, 1, 1]
+T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + p3[ax0, ax1, 0, 0])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 17 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 256], [1, 1, 256, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 17 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 256], [1, 1, 256, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 18 (workload key: ["7d44c6e3c81cd80f61ff2265b2bae89a", [1, 1024], [1000, 1024], [1, 1000], [1, 1000]]) ==========
-placeholder = PLACEHOLDER [1, 1024]
-placeholder = PLACEHOLDER [1000, 1024]
-T_matmul_NT(i, j) += (placeholder[i, k]*placeholder[j, k])
-placeholder = PLACEHOLDER [1, 1000]
-T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + placeholder[ax0, ax1])
-
-========== Task 19 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 512]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 512, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 18 (workload key: ["00a059b856ac30ac172b6252254479a6", [1, 1024], [1000, 1024], [1, 1000], [1, 1000]]) ==========
+p0 = PLACEHOLDER [1, 1024]
+p1 = PLACEHOLDER [1000, 1024]
+T_matmul_NT(i, j) += (p0[i, k]*p1[j, k])
+p2 = PLACEHOLDER [1, 1000]
+T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + p2[ax0, ax1])
+
+========== Task 19 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 512]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 512, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 20 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 7, 7, 1024], [3, 3, 1024, 1], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 1024]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 1024, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 1024]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 20 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 7, 7, 1024], [3, 3, 1024, 1], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 1024]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 1024, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 1024]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 21 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 112, 112, 32], [1, 1, 32, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
-placeholder = PLACEHOLDER [1, 112, 112, 32]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 32, 64]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 21 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 112, 112, 32], [1, 1, 32, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
+p0 = PLACEHOLDER [1, 112, 112, 32]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 32, 64]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
</pre></div>
</div>
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 31bb0ee18..cae2fdc76 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
@@ -502,277 +502,277 @@ The task scheduler will just optimize this objective.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Extract tasks...
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
-========== Task 0 (workload key: ["8654f16aeddf785bad9f028164b3a48d", [1, 56, 56, 64], [1, 1, 64, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 64, 64]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-
-========== Task 1 (workload key: ["c4500b4e2fd04e695c32d2f31bbdc14a", [1, 28, 28, 128], [4, 4, 128, 128], [1, 28, 28, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 0 (workload key: ["2d10de6646307f0e3e5cf4b31c20e69b", [1, 56, 56, 64], [1, 1, 64, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 64, 64]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+
+========== Task 1 (workload key: ["f19692ed81d032b1697c08adee62f9a5", [1, 28, 28, 128], [4, 4, 128, 128], [1, 28, 28, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*2) + eps), ((floormod(p, 14)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 128, 128]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 128, 128]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*14)*14) + (floordiv(h, 2)*14)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 28, 28, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+p3 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 2 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 56, 56, 64], [1, 1, 64, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 64, 128]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
+========== Task 2 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 56, 56, 64], [1, 1, 64, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 64, 128]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
-========== Task 3 (workload key: ["b8b52b9be9df6102466a22a014c44c1f", [1, 14, 14, 256], [4, 4, 256, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 3 (workload key: ["1097323f3970e5c881ad3a0028ca79cb", [1, 14, 14, 256], [4, 4, 256, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 49), ((floormod(floordiv(p, 7), 7)*2) + eps), ((floormod(p, 7)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 256, 256]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 256, 256]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*7)*7) + (floordiv(h, 2)*7)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 4 (workload key: ["e4cdf917b876dbdd64488c3818d9c141", [1, 28, 28, 128], [4, 4, 128, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 4 (workload key: ["0bcf718c0e6566bcd6c3b1437a3b6291", [1, 28, 28, 128], [4, 4, 128, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*2) + eps), ((floormod(p, 14)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 128, 128]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 128, 128]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*14)*14) + (floordiv(h, 2)*14)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 5 (workload key: ["d730bcd28f0920f6b97245e2a11bd8d6", [1, 7, 7, 512], [4, 4, 512, 512], [1, 7, 7, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 5 (workload key: ["d78e8eb6021c4cdda0ad7775d10f751a", [1, 7, 7, 512], [4, 4, 512, 512], [1, 7, 7, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 16), ((floormod(floordiv(p, 4), 4)*2) + eps), ((floormod(p, 4)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 512, 512]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 512, 512]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*4)*4) + (floordiv(h, 2)*4)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
+p2 = PLACEHOLDER [1, 7, 7, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
-========== Task 6 (workload key: ["b818b53148cd450f86569dfc3e04cb8a", [1, 56, 56, 64], [6, 6, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 6 (workload key: ["7c2a4f1f432f81c44985590780dfb52d", [1, 56, 56, 64], [6, 6, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*4) + eps), ((floormod(p, 14)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [6, 6, 64, 64]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [6, 6, 64, 64]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*14)*14) + (floordiv(h, 4)*14)) + floordiv(w, 4)), co]
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 7 (workload key: ["ad6cecbf5d85cb1cda3c2bb7af170211", [1, 7, 7, 512], [4, 4, 512, 512], [1, 7, 7, 512], [1, 1, 1, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 7 (workload key: ["a3df19e5b88592ef5a9ce584a1ca3010", [1, 7, 7, 512], [4, 4, 512, 512], [1, 7, 7, 512], [1, 1, 1, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 16), ((floormod(floordiv(p, 4), 4)*2) + eps), ((floormod(p, 4)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 512, 512]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 512, 512]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*4)*4) + (floordiv(h, 2)*4)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_multiply(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3]*placeholder[ax0, 0, 0, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 7, 7, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+p3 = PLACEHOLDER [1, 1, 1, 512]
+T_multiply(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3]*p3[ax0, 0, 0, ax3])
+p4 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + p4[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 8 (workload key: ["f3b6c10fcc6ce01ff01add933e4d21e9", [1, 14, 14, 256], [4, 4, 256, 256], [1, 14, 14, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 8 (workload key: ["64b7ce5264a64cb340d78b444b0325e6", [1, 14, 14, 256], [4, 4, 256, 256], [1, 14, 14, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 49), ((floormod(floordiv(p, 7), 7)*2) + eps), ((floormod(p, 7)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 256, 256]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 256, 256]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*7)*7) + (floordiv(h, 2)*7)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 14, 14, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+p3 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 9 (workload key: ["d7b65649a4dd54becea0a52aabbc5af5", [1, 1000], [1, 1000]]) ==========
-placeholder = PLACEHOLDER [1, 1000]
-T_softmax_maxelem(i0) max= placeholder[i0, k]
-T_softmax_exp(i0, i1) = tir.exp((placeholder[i0, i1] - T_softmax_maxelem[i0]))
+========== Task 9 (workload key: ["7d79c516e212fe1d73f5dbb90eaca2cf", [1, 1000], [1, 1000]]) ==========
+p0 = PLACEHOLDER [1, 1000]
+T_softmax_maxelem(i0) max= p0[i0, k]
+T_softmax_exp(i0, i1) = tir.exp((p0[i0, i1] - T_softmax_maxelem[i0]))
T_softmax_expsum(i0) += T_softmax_exp[i0, k]
T_softmax_norm(i0, i1) = (T_softmax_exp[i0, i1]/T_softmax_expsum[i0])
-========== Task 10 (workload key: ["69115f188984ae34ede37c3b8ca40b43", [1, 7, 7, 512], [1, 1, 1, 512]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-tensor(ax0, ax1, ax2, ax3) += placeholder[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
+========== Task 10 (workload key: ["be3babb9a46e32f66b717a3e2a2d522c", [1, 7, 7, 512], [1, 1, 1, 512]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+tensor(ax0, ax1, ax2, ax3) += p0[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
tensor(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3]/(float32((select((bool)1, ((ax1 + 1)*7), (((ax1 + 1)*7) + 1)) - (ax1*7)))*float32((select((bool)1, ((ax2 + 1)*7), (((ax2 + 1)*7) + 1)) - (ax2*7)))))
-========== Task 11 (workload key: ["3a69f9fbc63760d99e36b4c17b3bfc57", [1, 7, 7, 512], [4, 4, 512, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 11 (workload key: ["40b1cf1fd37b0ef111b3cc0247302508", [1, 7, 7, 512], [4, 4, 512, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 16), ((floormod(floordiv(p, 4), 4)*2) + eps), ((floormod(p, 4)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 512, 512]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 512, 512]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*4)*4) + (floordiv(h, 2)*4)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 12 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 28, 28, 128], [1, 1, 128, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 128, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-
-========== Task 13 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 14, 14, 256], [3, 3, 256, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 256, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 12 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 28, 28, 128], [1, 1, 128, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 128, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+
+========== Task 13 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 14, 14, 256], [3, 3, 256, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 256, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 14 (workload key: ["dac19035dd5fe9424ee8617421b9c817", [1, 28, 28, 128], [4, 4, 128, 128], [1, 28, 28, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 14 (workload key: ["25577781e50c611c2e45e73c1cb3a6ca", [1, 28, 28, 128], [4, 4, 128, 128], [1, 28, 28, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*2) + eps), ((floormod(p, 14)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 128, 128]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 128, 128]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*14)*14) + (floordiv(h, 2)*14)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
-========== Task 15 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 28, 28, 128], [3, 3, 128, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 128, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 28, 28, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+========== Task 15 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 28, 28, 128], [3, 3, 128, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 128, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 16 (workload key: ["1e3c4211ffd2f2db91078ae4d04b779d", [1, 56, 56, 64], [6, 6, 64, 64], [1, 56, 56, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 16 (workload key: ["6c4f6234946e16bcf9e48bdf289f9200", [1, 56, 56, 64], [6, 6, 64, 64], [1, 56, 56, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*4) + eps), ((floormod(p, 14)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [6, 6, 64, 64]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [6, 6, 64, 64]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*14)*14) + (floordiv(h, 4)*14)) + floordiv(w, 4)), co]
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 56, 56, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+p3 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 17 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 224, 224, 3], [7, 7, 3, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
-placeholder = PLACEHOLDER [1, 224, 224, 3]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 3) && (i1 < 227)) && (i2 >= 3)) && (i2 < 227)), placeholder[i0, (i1 - 3), (i2 - 3), i3], 0f)
-placeholder = PLACEHOLDER [7, 7, 3, 64]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 17 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 224, 224, 3], [7, 7, 3, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
+p0 = PLACEHOLDER [1, 224, 224, 3]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 3) && (i1 < 227)) && (i2 >= 3)) && (i2 < 227)), p0[i0, (i1 - 3), (i2 - 3), i3], 0f)
+p1 = PLACEHOLDER [7, 7, 3, 64]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 18 (workload key: ["3ea73fb9b0364374730d09e068821f95", [1, 56, 56, 64], [6, 6, 64, 64], [1, 56, 56, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 18 (workload key: ["10b8215aaf2e14d47d40b4093e6f41a0", [1, 56, 56, 64], [6, 6, 64, 64], [1, 56, 56, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 196), ((floormod(floordiv(p, 14), 14)*4) + eps), ((floormod(p, 14)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [6, 6, 64, 64]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [6, 6, 64, 64]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*14)*14) + (floordiv(h, 4)*14)) + floordiv(w, 4)), co]
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
+p2 = PLACEHOLDER [1, 56, 56, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
-========== Task 19 (workload key: ["d374e472bd9d8164892b9e28a0a8cb59", [1, 14, 14, 256], [4, 4, 256, 256], [1, 14, 14, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 19 (workload key: ["7f3fee61bc3c2604395f5d343b840b7c", [1, 14, 14, 256], [4, 4, 256, 256], [1, 14, 14, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 49), ((floormod(floordiv(p, 7), 7)*2) + eps), ((floormod(p, 7)*2) + nu), ci]
B(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 4) == 2)), ..(OMITTED).. ormod(i, 4) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [4, 4, 256, 256]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [4, 4, 256, 256]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 4) == 3) && (floormod(j, 2) == 1)), 1f, select(((floormod(i, 4) == 3) && (floormod(j, 2) == 0)), ..(OMITTED).. ct(((floormod(i, 4) == 0) && (floormod(j, 2) == 1)), 0f, select(((floormod(i, 4) == 0) && (floormod(j, 2) == 0)), 1f, 0f))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 2), floormod(w, 2), ((((n*7)*7) + (floordiv(h, 2)*7)) + floordiv(w, 2)), co]
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
+p2 = PLACEHOLDER [1, 14, 14, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
-========== Task 20 (workload key: ["64b98c71af70a904fdbb81d7d4188d84", [1, 112, 112, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 112, 112, 64]
-pad_temp(ax0, ax1, ax2, ax3) = tir.if_then_else(((((ax1 >= 1) && (ax1 < 113)) && (ax2 >= 1)) && (ax2 < 113)), placeholder[ax0, (ax1 - 1), (ax2 - 1), ax3], -3.40282e+38f)
+========== Task 20 (workload key: ["affd3c4a65f665e451a06d65bf32750d", [1, 112, 112, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 112, 112, 64]
+pad_temp(ax0, ax1, ax2, ax3) = tir.if_then_else(((((ax1 >= 1) && (ax1 < 113)) && (ax2 >= 1)) && (ax2 < 113)), p0[ax0, (ax1 - 1), (ax2 - 1), ax3], -3.40282e+38f)
tensor(ax0, ax1, ax2, ax3) max= pad_temp[ax0, ((ax1*2) + rv0), ((ax2*2) + rv1), ax3]
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p1 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3] + p1[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 21 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 14, 14, 256], [1, 1, 256, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-
-========== Task 22 (workload key: ["7d44c6e3c81cd80f61ff2265b2bae89a", [1, 512], [1000, 512], [1, 1000], [1, 1000]]) ==========
-placeholder = PLACEHOLDER [1, 512]
-placeholder = PLACEHOLDER [1000, 512]
-T_matmul_NT(i, j) += (placeholder[i, k]*placeholder[j, k])
-placeholder = PLACEHOLDER [1, 1000]
-T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + placeholder[ax0, ax1])
-
-========== Task 23 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 56, 56, 64], [3, 3, 64, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 64, 128]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 21 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 14, 14, 256], [1, 1, 256, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+
+========== Task 22 (workload key: ["00a059b856ac30ac172b6252254479a6", [1, 512], [1000, 512], [1, 1000], [1, 1000]]) ==========
+p0 = PLACEHOLDER [1, 512]
+p1 = PLACEHOLDER [1000, 512]
+T_matmul_NT(i, j) += (p0[i, k]*p1[j, k])
+p2 = PLACEHOLDER [1, 1000]
+T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + p2[ax0, ax1])
+
+========== Task 23 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 56, 56, 64], [3, 3, 64, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 64, 128]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
</pre></div>
</div>
@@ -906,7 +906,7 @@ 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)
- 10.0240 10.0604 10.0697 9.9419 0.0582
+ 8.1562 8.1569 8.1608 8.1510 0.0040
</pre></div>
</div>
</div>
diff --git a/docs/how_to/tune_with_autoscheduler/tune_network_mali.html b/docs/how_to/tune_with_autoscheduler/tune_network_mali.html
index eaa065a0e..e4ce92fe4 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_mali.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_mali.html
@@ -517,196 +517,196 @@ The task scheduler will just optimize this objective.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Extract tasks...
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
-========== Task 0 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 1024], [1, 1, 1024, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 1024]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 1024, 1024]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 1024]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 0 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 1024], [1, 1, 1024, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 1024]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 1024, 1024]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 1024]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 1 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 256], [1, 1, 256, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 1 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 256], [1, 1, 256, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 2 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 256]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 256, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 2 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 256]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 256, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 3 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 128], [1, 1, 128, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 128, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 3 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 128], [1, 1, 128, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 128, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 4 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 512], [1, 1, 512, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 1024]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 1024]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 4 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 512], [1, 1, 512, 1024], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 1024]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 1024]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 5 (workload key: ["69115f188984ae34ede37c3b8ca40b43", [1, 7, 7, 1024], [1, 1, 1, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 1024]
-tensor(ax0, ax1, ax2, ax3) += placeholder[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
+========== Task 5 (workload key: ["be3babb9a46e32f66b717a3e2a2d522c", [1, 7, 7, 1024], [1, 1, 1, 1024]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 1024]
+tensor(ax0, ax1, ax2, ax3) += p0[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
tensor(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3]/(float32((select((bool)1, ((ax1 + 1)*7), (((ax1 + 1)*7) + 1)) - (ax1*7)))*float32((select((bool)1, ((ax2 + 1)*7), (((ax2 + 1)*7) + 1)) - (ax2*7)))))
-========== Task 6 (workload key: ["d7b65649a4dd54becea0a52aabbc5af5", [1, 1000], [1, 1000]]) ==========
-placeholder = PLACEHOLDER [1, 1000]
-T_softmax_maxelem(i0) max= placeholder[i0, k]
-T_softmax_exp(i0, i1) = tir.exp((placeholder[i0, i1] - T_softmax_maxelem[i0]))
+========== Task 6 (workload key: ["7d79c516e212fe1d73f5dbb90eaca2cf", [1, 1000], [1, 1000]]) ==========
+p0 = PLACEHOLDER [1, 1000]
+T_softmax_maxelem(i0) max= p0[i0, k]
+T_softmax_exp(i0, i1) = tir.exp((p0[i0, i1] - T_softmax_maxelem[i0]))
T_softmax_expsum(i0) += T_softmax_exp[i0, k]
T_softmax_norm(i0, i1) = (T_softmax_exp[i0, i1]/T_softmax_expsum[i0])
-========== Task 7 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 128], [1, 1, 128, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 128]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 128, 128]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 7 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 128], [1, 1, 128, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 128]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 128, 128]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 8 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 7, 7, 1024], [3, 3, 1024, 1], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 1024]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 1024, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 1024]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 8 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 7, 7, 1024], [3, 3, 1024, 1], [1, 1, 1, 1024], [1, 7, 7, 1024]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 1024]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 1024, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 1024]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 9 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 64], [1, 1, 64, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 64, 128]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 9 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 64], [1, 1, 64, 128], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 64, 128]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 10 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 512]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 512, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 10 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 512]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 512, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 11 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 256], [1, 1, 256, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 11 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 256], [1, 1, 256, 256], [1, 1, 1, 256], [1, 28, 28, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 12 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 128]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 128, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 12 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 56, 56, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 128]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 128, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 13 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 112, 112, 64], [3, 3, 64, 1], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 112, 112, 64]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 64, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 13 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 112, 112, 64], [3, 3, 64, 1], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 112, 112, 64]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 64, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 14 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 112, 112, 32], [3, 3, 32, 1], [1, 1, 1, 32], [1, 112, 112, 32]]) ==========
-placeholder = PLACEHOLDER [1, 112, 112, 32]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 32, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 32]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 14 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 112, 112, 32], [3, 3, 32, 1], [1, 1, 1, 32], [1, 112, 112, 32]]) ==========
+p0 = PLACEHOLDER [1, 112, 112, 32]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 113)) && (i2 >= 1)) && (i2 < 113)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 32, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 32]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 15 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 128]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 128, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 15 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 56, 56, 128], [3, 3, 128, 1], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 128]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 128, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 16 (workload key: ["2ca148ecea6508ce625f85719021344f", [1, 224, 224, 3], [3, 3, 3, 32], [1, 112, 1, 1], [1, 112, 1, 1], [1, 112, 112, 32]]) ==========
-placeholder = PLACEHOLDER [1, 224, 224, 3]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 225)) && (i2 >= 1)) && (i2 < 225)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 3, 32]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 112, 1, 1]
-T_multiply(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3]*placeholder[ax0, ax1, 0, 0])
-placeholder = PLACEHOLDER [1, 112, 1, 1]
-T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, 0, 0])
+========== Task 16 (workload key: ["ad24d4d2f83975ff580a4833fbf3ef94", [1, 224, 224, 3], [3, 3, 3, 32], [1, 112, 1, 1], [1, 112, 1, 1], [1, 112, 112, 32]]) ==========
+p0 = PLACEHOLDER [1, 224, 224, 3]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 225)) && (i2 >= 1)) && (i2 < 225)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 3, 32]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 112, 1, 1]
+T_multiply(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3]*p2[ax0, ax1, 0, 0])
+p3 = PLACEHOLDER [1, 112, 1, 1]
+T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + p3[ax0, ax1, 0, 0])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 17 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 512], [1, 1, 512, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 17 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 512], [1, 1, 512, 512], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 18 (workload key: ["7d44c6e3c81cd80f61ff2265b2bae89a", [1, 1024], [1000, 1024], [1, 1000], [1, 1000]]) ==========
-placeholder = PLACEHOLDER [1, 1024]
-placeholder = PLACEHOLDER [1000, 1024]
-T_matmul_NT(i, j) += (placeholder[i, k]*placeholder[j, k])
-placeholder = PLACEHOLDER [1, 1000]
-T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + placeholder[ax0, ax1])
-
-========== Task 19 (workload key: ["06fce76bd84cb904eee50b905ca9449a", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 512]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 512, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 18 (workload key: ["00a059b856ac30ac172b6252254479a6", [1, 1024], [1000, 1024], [1, 1000], [1, 1000]]) ==========
+p0 = PLACEHOLDER [1, 1024]
+p1 = PLACEHOLDER [1000, 1024]
+T_matmul_NT(i, j) += (p0[i, k]*p1[j, k])
+p2 = PLACEHOLDER [1, 1000]
+T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + p2[ax0, ax1])
+
+========== Task 19 (workload key: ["98cde4888c94ec7beaa9972f806856d0", [1, 14, 14, 512], [3, 3, 512, 1], [1, 1, 1, 512], [1, 14, 14, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 512]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 512, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, (i + di), (j + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 20 (workload key: ["c87ba68bc180312f5716af09a77ca15b", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 256]
-PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 256, 1]
-DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*placeholder[di, dj, c, 0])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 20 (workload key: ["88a2e34d300a6ccfcf0228f0b90f13ec", [1, 28, 28, 256], [3, 3, 256, 1], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 256]
+PaddedInput(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 256, 1]
+DepthwiseConv2d(b, i, j, c) += (PaddedInput[b, ((i*2) + di), ((j*2) + dj), c]*p1[di, dj, c, 0])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (DepthwiseConv2d[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 21 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 112, 112, 32], [1, 1, 32, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
-placeholder = PLACEHOLDER [1, 112, 112, 32]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 32, 64]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 21 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 112, 112, 32], [1, 1, 32, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
+p0 = PLACEHOLDER [1, 112, 112, 32]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 32, 64]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
</pre></div>
</div>
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 77b4bf26d..df0ba7a8a 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
@@ -525,262 +525,262 @@ The task scheduler will just optimize this objective.</p>
Extract tasks...
/workspace/python/tvm/driver/build_module.py:267: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
-========== Task 0 (workload key: ["8654f16aeddf785bad9f028164b3a48d", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 64, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-
-========== Task 1 (workload key: ["12cb81d4ad0a81be02dedf09d1ac8391", [1, 14, 14, 256], [1, 1, 256, 1024], [1, 14, 14, 1024], [1, 14, 14, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 1024]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 14, 14, 1024]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
-========== Task 2 (workload key: ["b9a4f9bd1416ba25810cb3de27628ace", [1, 14, 14, 256], [1, 1, 256, 1024], [1, 14, 14, 1024], [1, 1, 1, 1024], [1, 14, 14, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 1024]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 14, 14, 1024]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 1024]
-T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 0 (workload key: ["2d10de6646307f0e3e5cf4b31c20e69b", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 64, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+
+========== Task 1 (workload key: ["3060808fc5c74e18b1276729071fbae0", [1, 14, 14, 256], [1, 1, 256, 1024], [1, 14, 14, 1024], [1, 14, 14, 1024]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 1024]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 14, 14, 1024]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+========== Task 2 (workload key: ["76afb7bf408a1ffa0b8b7bc09d077dc3", [1, 14, 14, 256], [1, 1, 256, 1024], [1, 14, 14, 1024], [1, 1, 1, 1024], [1, 14, 14, 1024]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 1024]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 14, 14, 1024]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+p3 = PLACEHOLDER [1, 1, 1, 1024]
+T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 3 (workload key: ["56af7508fbcdf6d851892b1e8434667b", [1, 14, 14, 1024], [1, 1, 1024, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 1024]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 1024, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 3 (workload key: ["2beb39e9afe4c74822fffbcbb8533595", [1, 14, 14, 1024], [1, 1, 1024, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 1024]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 1024, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 4 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 56, 56, 256], [1, 1, 256, 512], [1, 28, 28, 512]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
+========== Task 4 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 56, 56, 256], [1, 1, 256, 512], [1, 28, 28, 512]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
-========== Task 5 (workload key: ["c68f92478eb18145106184c587d212b6", [1, 14, 14, 256], [6, 6, 256, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 256]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 5 (workload key: ["38552500208b25b4035682b0e93cbce3", [1, 14, 14, 256], [6, 6, 256, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 256]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 15)) && (i2 >= 1)) && (i2 < 15)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 16), ((floormod(floordiv(p, 4), 4)*4) + eps), ((floormod(p, 4)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [6, 6, 256, 256]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [6, 6, 256, 256]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*4)*4) + (floordiv(h, 4)*4)) + floordiv(w, 4)), co]
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 6 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 14, 14, 1024], [1, 1, 1024, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 1024]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 1024, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 6 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 14, 14, 1024], [1, 1, 1024, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 1024]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 1024, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 7 (workload key: ["12cb81d4ad0a81be02dedf09d1ac8391", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256], [1, 56, 56, 256]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 64, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 56, 56, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
-========== Task 8 (workload key: ["ecec634b4882c5731f86cce3109db636", [1, 28, 28, 128], [6, 6, 128, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
+========== Task 7 (workload key: ["3060808fc5c74e18b1276729071fbae0", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256], [1, 56, 56, 256]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 64, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 56, 56, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+========== Task 8 (workload key: ["cfd09cf1ca9e943f0ee12a18813a5c75", [1, 28, 28, 128], [6, 6, 128, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+data_pad(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 29)) && (i2 >= 1)) && (i2 < 29)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
input_tile(eps, nu, p, ci) = data_pad[floordiv(p, 49), ((floormod(floordiv(p, 7), 7)*4) + eps), ((floormod(p, 7)*4) + nu), ci]
B(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 6) == 5)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 6) == 4)), ..(OMITTED).. (floormod(j, 6) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 6) == 0)), 1f, 0f))))))))))))))))))))))))))))))))))))
data_pack(eps, nu, p, ci) += ((input_tile[r_a, r_b, p, ci]*B[r_a, eps])*B[r_b, nu])
-placeholder = PLACEHOLDER [6, 6, 128, 128]
-bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*placeholder[eps, nu, co, ci])
+p1 = PLACEHOLDER [6, 6, 128, 128]
+bgemm(eps, nu, p, co) += (data_pack[eps, nu, p, ci]*p1[eps, nu, co, ci])
A(i, j) = select(((floormod(i, 6) == 5) && (floormod(j, 4) == 3)), 1f, select(((floormod(i, 6) == 5) && (floormod(j, 4) == 2)), ..(OMITTED).. 6) == 0) && (floormod(j, 4) == 1)), 0f, select(((floormod(i, 6) == 0) && (floormod(j, 4) == 0)), 1f, 0f))))))))))))))))))))))))
inverse(vh, vw, p, co) += ((bgemm[r_a, r_b, p, co]*A[r_a, vh])*A[r_b, vw])
conv2d_winograd(n, h, w, co) = inverse[floormod(h, 4), floormod(w, 4), ((((n*7)*7) + (floordiv(h, 4)*7)) + floordiv(w, 4)), co]
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_winograd[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 9 (workload key: ["12cb81d4ad0a81be02dedf09d1ac8391", [1, 28, 28, 128], [1, 1, 128, 512], [1, 28, 28, 512], [1, 28, 28, 512]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 128, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 28, 28, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
-========== Task 10 (workload key: ["d7b65649a4dd54becea0a52aabbc5af5", [1, 1000], [1, 1000]]) ==========
-placeholder = PLACEHOLDER [1, 1000]
-T_softmax_maxelem(i0) max= placeholder[i0, k]
-T_softmax_exp(i0, i1) = tir.exp((placeholder[i0, i1] - T_softmax_maxelem[i0]))
+========== Task 9 (workload key: ["3060808fc5c74e18b1276729071fbae0", [1, 28, 28, 128], [1, 1, 128, 512], [1, 28, 28, 512], [1, 28, 28, 512]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 128, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 28, 28, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+========== Task 10 (workload key: ["7d79c516e212fe1d73f5dbb90eaca2cf", [1, 1000], [1, 1000]]) ==========
+p0 = PLACEHOLDER [1, 1000]
+T_softmax_maxelem(i0) max= p0[i0, k]
+T_softmax_exp(i0, i1) = tir.exp((p0[i0, i1] - T_softmax_maxelem[i0]))
T_softmax_expsum(i0) += T_softmax_exp[i0, k]
T_softmax_norm(i0, i1) = (T_softmax_exp[i0, i1]/T_softmax_expsum[i0])
-========== Task 11 (workload key: ["69115f188984ae34ede37c3b8ca40b43", [1, 7, 7, 2048], [1, 1, 1, 2048]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 2048]
-tensor(ax0, ax1, ax2, ax3) += placeholder[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
+========== Task 11 (workload key: ["be3babb9a46e32f66b717a3e2a2d522c", [1, 7, 7, 2048], [1, 1, 1, 2048]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 2048]
+tensor(ax0, ax1, ax2, ax3) += p0[ax0, ((ax1*7) + rv0), ((ax2*7) + rv1), ax3]
tensor(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3]/(float32((select((bool)1, ((ax1 + 1)*7), (((ax1 + 1)*7) + 1)) - (ax1*7)))*float32((select((bool)1, ((ax2 + 1)*7), (((ax2 + 1)*7) + 1)) - (ax2*7)))))
-========== Task 12 (workload key: ["12cb81d4ad0a81be02dedf09d1ac8391", [1, 7, 7, 512], [1, 1, 512, 2048], [1, 7, 7, 2048], [1, 7, 7, 2048]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 2048]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 7, 7, 2048]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-
-========== Task 13 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 28, 28, 512], [1, 1, 512, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 128]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 12 (workload key: ["3060808fc5c74e18b1276729071fbae0", [1, 7, 7, 512], [1, 1, 512, 2048], [1, 7, 7, 2048], [1, 7, 7, 2048]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 2048]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 7, 7, 2048]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+
+========== Task 13 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 28, 28, 512], [1, 1, 512, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 128]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 14 (workload key: ["56af7508fbcdf6d851892b1e8434667b", [1, 28, 28, 512], [1, 1, 512, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 14 (workload key: ["2beb39e9afe4c74822fffbcbb8533595", [1, 28, 28, 512], [1, 1, 512, 256], [1, 1, 1, 256], [1, 14, 14, 256]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 15 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 28, 28, 512], [1, 1, 512, 1024], [1, 14, 14, 1024]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 1024]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-
-========== Task 16 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 7, 7, 2048], [1, 1, 2048, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 2048]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 2048, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 15 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 28, 28, 512], [1, 1, 512, 1024], [1, 14, 14, 1024]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 1024]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+
+========== Task 16 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 7, 7, 2048], [1, 1, 2048, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 2048]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 2048, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 17 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 256], [1, 1, 256, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 64]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 17 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 256], [1, 1, 256, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 64]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 18 (workload key: ["b9a4f9bd1416ba25810cb3de27628ace", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256], [1, 1, 1, 256], [1, 56, 56, 256]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 64, 256]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 56, 56, 256]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 256]
-T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 18 (workload key: ["76afb7bf408a1ffa0b8b7bc09d077dc3", [1, 56, 56, 64], [1, 1, 64, 256], [1, 56, 56, 256], [1, 1, 1, 256], [1, 56, 56, 256]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 64, 256]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 56, 56, 256]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+p3 = PLACEHOLDER [1, 1, 1, 256]
+T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 19 (workload key: ["b9a4f9bd1416ba25810cb3de27628ace", [1, 28, 28, 128], [1, 1, 128, 512], [1, 28, 28, 512], [1, 1, 1, 512], [1, 28, 28, 512]]) ==========
-placeholder = PLACEHOLDER [1, 28, 28, 128]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 128, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 28, 28, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 19 (workload key: ["76afb7bf408a1ffa0b8b7bc09d077dc3", [1, 28, 28, 128], [1, 1, 128, 512], [1, 28, 28, 512], [1, 1, 1, 512], [1, 28, 28, 512]]) ==========
+p0 = PLACEHOLDER [1, 28, 28, 128]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 128, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 28, 28, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+p3 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3] + p3[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 20 (workload key: ["86551f1a74663d3ceafd5884659d3478", [1, 7, 7, 512], [3, 3, 512, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 512, 512]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 512]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 20 (workload key: ["d37380659057397544e056461ea3bad3", [1, 7, 7, 512], [3, 3, 512, 512], [1, 1, 1, 512], [1, 7, 7, 512]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 8)) && (i2 >= 1)) && (i2 < 8)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 512, 512]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 512]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 21 (workload key: ["86551f1a74663d3ceafd5884659d3478", [1, 56, 56, 64], [3, 3, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), placeholder[i0, (i1 - 1), (i2 - 1), i3], 0f)
-placeholder = PLACEHOLDER [3, 3, 64, 64]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 21 (workload key: ["d37380659057397544e056461ea3bad3", [1, 56, 56, 64], [3, 3, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 1) && (i1 < 57)) && (i2 >= 1)) && (i2 < 57)), p0[i0, (i1 - 1), (i2 - 1), i3], 0f)
+p1 = PLACEHOLDER [3, 3, 64, 64]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 22 (workload key: ["56af7508fbcdf6d851892b1e8434667b", [1, 56, 56, 256], [1, 1, 256, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 256]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 256, 128]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 128]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 22 (workload key: ["2beb39e9afe4c74822fffbcbb8533595", [1, 56, 56, 256], [1, 1, 256, 128], [1, 1, 1, 128], [1, 28, 28, 128]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 256]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 256, 128]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 128]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 23 (workload key: ["ff5ea7f814e5c497bb685e7385cf7159", [1, 7, 7, 512], [1, 1, 512, 2048], [1, 7, 7, 2048], [1, 1, 1, 2048], [1, 1, 1, 2048], [1, 7, 7, 2048]]) ==========
-placeholder = PLACEHOLDER [1, 7, 7, 512]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 512, 2048]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 7, 7, 2048]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, ax1, ax2, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 2048]
-T_multiply(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3]*placeholder[ax0, 0, 0, ax3])
-placeholder = PLACEHOLDER [1, 1, 1, 2048]
-T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 23 (workload key: ["f07e228ef5f642b386d23a62df615e7b", [1, 7, 7, 512], [1, 1, 512, 2048], [1, 7, 7, 2048], [1, 1, 1, 2048], [1, 1, 1, 2048], [1, 7, 7, 2048]]) ==========
+p0 = PLACEHOLDER [1, 7, 7, 512]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 512, 2048]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 7, 7, 2048]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, ax1, ax2, ax3])
+p3 = PLACEHOLDER [1, 1, 1, 2048]
+T_multiply(ax0, ax1, ax2, ax3) = (T_add[ax0, ax1, ax2, ax3]*p3[ax0, 0, 0, ax3])
+p4 = PLACEHOLDER [1, 1, 1, 2048]
+T_add(ax0, ax1, ax2, ax3) = (T_multiply[ax0, ax1, ax2, ax3] + p4[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 24 (workload key: ["96daaa9daa1b41bc383b7c05ce8b58de", [1, 224, 224, 3], [7, 7, 3, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
-placeholder = PLACEHOLDER [1, 224, 224, 3]
-pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 3) && (i1 < 227)) && (i2 >= 3)) && (i2 < 227)), placeholder[i0, (i1 - 3), (i2 - 3), i3], 0f)
-placeholder = PLACEHOLDER [7, 7, 3, 64]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 24 (workload key: ["07f9fcad27bdd3233f86fe35a5185d33", [1, 224, 224, 3], [7, 7, 3, 64], [1, 1, 1, 64], [1, 112, 112, 64]]) ==========
+p0 = PLACEHOLDER [1, 224, 224, 3]
+pad_temp(i0, i1, i2, i3) = tir.if_then_else(((((i1 >= 3) && (i1 < 227)) && (i2 >= 3)) && (i2 < 227)), p0[i0, (i1 - 3), (i2 - 3), i3], 0f)
+p1 = PLACEHOLDER [7, 7, 3, 64]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 25 (workload key: ["06f578e6519a86e85028eecf4de64b25", [1, 14, 14, 1024], [1, 1, 1024, 2048], [1, 7, 7, 2048]]) ==========
-placeholder = PLACEHOLDER [1, 14, 14, 1024]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 1024, 2048]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*placeholder[ry, rx, rc, ff])
-
-========== Task 26 (workload key: ["7d44c6e3c81cd80f61ff2265b2bae89a", [1, 2048], [1000, 2048], [1, 1000], [1, 1000]]) ==========
-placeholder = PLACEHOLDER [1, 2048]
-placeholder = PLACEHOLDER [1000, 2048]
-T_matmul_NT(i, j) += (placeholder[i, k]*placeholder[j, k])
-placeholder = PLACEHOLDER [1, 1000]
-T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + placeholder[ax0, ax1])
-
-========== Task 27 (workload key: ["64b98c71af70a904fdbb81d7d4188d84", [1, 112, 112, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 112, 112, 64]
-pad_temp(ax0, ax1, ax2, ax3) = tir.if_then_else(((((ax1 >= 1) && (ax1 < 113)) && (ax2 >= 1)) && (ax2 < 113)), placeholder[ax0, (ax1 - 1), (ax2 - 1), ax3], -3.40282e+38f)
+========== Task 25 (workload key: ["0fad1b42d0d33418e0a8d15d3bbad3c9", [1, 14, 14, 1024], [1, 1, 1024, 2048], [1, 7, 7, 2048]]) ==========
+p0 = PLACEHOLDER [1, 14, 14, 1024]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 1024, 2048]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, ((yy*2) + ry), ((xx*2) + rx), rc]*p1[ry, rx, rc, ff])
+
+========== Task 26 (workload key: ["00a059b856ac30ac172b6252254479a6", [1, 2048], [1000, 2048], [1, 1000], [1, 1000]]) ==========
+p0 = PLACEHOLDER [1, 2048]
+p1 = PLACEHOLDER [1000, 2048]
+T_matmul_NT(i, j) += (p0[i, k]*p1[j, k])
+p2 = PLACEHOLDER [1, 1000]
+T_add(ax0, ax1) = (T_matmul_NT[ax0, ax1] + p2[ax0, ax1])
+
+========== Task 27 (workload key: ["affd3c4a65f665e451a06d65bf32750d", [1, 112, 112, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 112, 112, 64]
+pad_temp(ax0, ax1, ax2, ax3) = tir.if_then_else(((((ax1 >= 1) && (ax1 < 113)) && (ax2 >= 1)) && (ax2 < 113)), p0[ax0, (ax1 - 1), (ax2 - 1), ax3], -3.40282e+38f)
tensor(ax0, ax1, ax2, ax3) max= pad_temp[ax0, ((ax1*2) + rv0), ((ax2*2) + rv1), ax3]
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+p1 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (tensor[ax0, ax1, ax2, ax3] + p1[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
-========== Task 28 (workload key: ["1037be767e8e18197e87653d81c34558", [1, 56, 56, 64], [1, 1, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
-placeholder = PLACEHOLDER [1, 56, 56, 64]
-pad_temp(i0, i1, i2, i3) = placeholder[i0, i1, i2, i3]
-placeholder = PLACEHOLDER [1, 1, 64, 64]
-conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*placeholder[ry, rx, rc, ff])
-placeholder = PLACEHOLDER [1, 1, 1, 64]
-T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + placeholder[ax0, 0, 0, ax3])
+========== Task 28 (workload key: ["6d628209072e3e3dd8f49359935acea6", [1, 56, 56, 64], [1, 1, 64, 64], [1, 1, 1, 64], [1, 56, 56, 64]]) ==========
+p0 = PLACEHOLDER [1, 56, 56, 64]
+pad_temp(i0, i1, i2, i3) = p0[i0, i1, i2, i3]
+p1 = PLACEHOLDER [1, 1, 64, 64]
+conv2d_nhwc(nn, yy, xx, ff) += (pad_temp[nn, (yy + ry), (xx + rx), rc]*p1[ry, rx, rc, ff])
+p2 = PLACEHOLDER [1, 1, 1, 64]
+T_add(ax0, ax1, ax2, ax3) = (conv2d_nhwc[ax0, ax1, ax2, ax3] + p2[ax0, 0, 0, ax3])
T_relu(ax0, ax1, ax2, ax3) = max(T_add[ax0, ax1, ax2, ax3], 0f)
</pre></div>
</div>
@@ -925,7 +925,7 @@ 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)
- 757.9706 757.1987 759.9952 756.7179 1.4450
+ 760.1457 760.9216 762.4794 757.0362 2.2889
</pre></div>
</div>
</div>
@@ -947,7 +947,7 @@ 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 23.325 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 24.369 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 6f7c9cfef..97f36566e 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
@@ -625,792 +625,409 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [65536], []),
compute: Buffer(compute_2: Pointer(float32), float32, [65536], [])}
buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute}
- preflattened_buffer_map = {placeholder_6: placeholder_15: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_7: placeholder_16: Buffer(placeholder_12, int32, [4916], []), placeholder_5: placeholder_17: Buffer(placeholder_10, float32, [128, 256], []), placeholder_9: placeholder_18: Buffer(placeholder_14, float32, [128, 512], []), placeholder_8: placeholder_19: Buffer(placeholder_13, int32, [33], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], [])} {
- for (i0.outer: int32, 0, 8) "parallel" {
- allocate(compute_4: Pointer(global float32), float32, [512]), storage_scope = global;
- for (i1.outer: int32, 0, 16) {
- for (nb_j.inner: int32, 0, 2) {
- let cse_var_2: int32 = (nb_j.inner*16)
- let cse_var_1: int32 = ((i1.outer*2) + nb_j.inner)
- {
- compute_5: Buffer(compute_4, float32, [512], [])[cse_var_2] = 0f32
- compute_5[(cse_var_2 + 1)] = 0f32
- compute_5[(cse_var_2 + 2)] = 0f32
- compute_5[(cse_var_2 + 3)] = 0f32
- compute_5[(cse_var_2 + 4)] = 0f32
- compute_5[(cse_var_2 + 5)] = 0f32
- compute_5[(cse_var_2 + 6)] = 0f32
- compute_5[(cse_var_2 + 7)] = 0f32
- compute_5[(cse_var_2 + 8)] = 0f32
- compute_5[(cse_var_2 + 9)] = 0f32
- compute_5[(cse_var_2 + 10)] = 0f32
- compute_5[(cse_var_2 + 11)] = 0f32
- compute_5[(cse_var_2 + 12)] = 0f32
- compute_5[(cse_var_2 + 13)] = 0f32
- compute_5[(cse_var_2 + 14)] = 0f32
- compute_5[(cse_var_2 + 15)] = 0f32
- compute_5[(cse_var_2 + 32)] = 0f32
- compute_5[(cse_var_2 + 33)] = 0f32
- compute_5[(cse_var_2 + 34)] = 0f32
- compute_5[(cse_var_2 + 35)] = 0f32
- compute_5[(cse_var_2 + 36)] = 0f32
- compute_5[(cse_var_2 + 37)] = 0f32
- compute_5[(cse_var_2 + 38)] = 0f32
- compute_5[(cse_var_2 + 39)] = 0f32
- compute_5[(cse_var_2 + 40)] = 0f32
- compute_5[(cse_var_2 + 41)] = 0f32
- compute_5[(cse_var_2 + 42)] = 0f32
- compute_5[(cse_var_2 + 43)] = 0f32
- compute_5[(cse_var_2 + 44)] = 0f32
- compute_5[(cse_var_2 + 45)] = 0f32
- compute_5[(cse_var_2 + 46)] = 0f32
- compute_5[(cse_var_2 + 47)] = 0f32
- compute_5[(cse_var_2 + 64)] = 0f32
- compute_5[(cse_var_2 + 65)] = 0f32
- compute_5[(cse_var_2 + 66)] = 0f32
- compute_5[(cse_var_2 + 67)] = 0f32
- compute_5[(cse_var_2 + 68)] = 0f32
- compute_5[(cse_var_2 + 69)] = 0f32
- compute_5[(cse_var_2 + 70)] = 0f32
- compute_5[(cse_var_2 + 71)] = 0f32
- compute_5[(cse_var_2 + 72)] = 0f32
- compute_5[(cse_var_2 + 73)] = 0f32
- compute_5[(cse_var_2 + 74)] = 0f32
- compute_5[(cse_var_2 + 75)] = 0f32
- compute_5[(cse_var_2 + 76)] = 0f32
- compute_5[(cse_var_2 + 77)] = 0f32
- compute_5[(cse_var_2 + 78)] = 0f32
- compute_5[(cse_var_2 + 79)] = 0f32
- compute_5[(cse_var_2 + 96)] = 0f32
- compute_5[(cse_var_2 + 97)] = 0f32
- compute_5[(cse_var_2 + 98)] = 0f32
- compute_5[(cse_var_2 + 99)] = 0f32
- compute_5[(cse_var_2 + 100)] = 0f32
- compute_5[(cse_var_2 + 101)] = 0f32
- compute_5[(cse_var_2 + 102)] = 0f32
- compute_5[(cse_var_2 + 103)] = 0f32
- compute_5[(cse_var_2 + 104)] = 0f32
- compute_5[(cse_var_2 + 105)] = 0f32
- compute_5[(cse_var_2 + 106)] = 0f32
- compute_5[(cse_var_2 + 107)] = 0f32
- compute_5[(cse_var_2 + 108)] = 0f32
- compute_5[(cse_var_2 + 109)] = 0f32
- compute_5[(cse_var_2 + 110)] = 0f32
- compute_5[(cse_var_2 + 111)] = 0f32
- compute_5[(cse_var_2 + 128)] = 0f32
- compute_5[(cse_var_2 + 129)] = 0f32
- compute_5[(cse_var_2 + 130)] = 0f32
- compute_5[(cse_var_2 + 131)] = 0f32
- compute_5[(cse_var_2 + 132)] = 0f32
- compute_5[(cse_var_2 + 133)] = 0f32
- compute_5[(cse_var_2 + 134)] = 0f32
- compute_5[(cse_var_2 + 135)] = 0f32
- compute_5[(cse_var_2 + 136)] = 0f32
- compute_5[(cse_var_2 + 137)] = 0f32
- compute_5[(cse_var_2 + 138)] = 0f32
- compute_5[(cse_var_2 + 139)] = 0f32
- compute_5[(cse_var_2 + 140)] = 0f32
- compute_5[(cse_var_2 + 141)] = 0f32
- compute_5[(cse_var_2 + 142)] = 0f32
- compute_5[(cse_var_2 + 143)] = 0f32
- compute_5[(cse_var_2 + 160)] = 0f32
- compute_5[(cse_var_2 + 161)] = 0f32
- compute_5[(cse_var_2 + 162)] = 0f32
- compute_5[(cse_var_2 + 163)] = 0f32
- compute_5[(cse_var_2 + 164)] = 0f32
- compute_5[(cse_var_2 + 165)] = 0f32
- compute_5[(cse_var_2 + 166)] = 0f32
- compute_5[(cse_var_2 + 167)] = 0f32
- compute_5[(cse_var_2 + 168)] = 0f32
- compute_5[(cse_var_2 + 169)] = 0f32
- compute_5[(cse_var_2 + 170)] = 0f32
- compute_5[(cse_var_2 + 171)] = 0f32
- compute_5[(cse_var_2 + 172)] = 0f32
- compute_5[(cse_var_2 + 173)] = 0f32
- compute_5[(cse_var_2 + 174)] = 0f32
- compute_5[(cse_var_2 + 175)] = 0f32
- compute_5[(cse_var_2 + 192)] = 0f32
- compute_5[(cse_var_2 + 193)] = 0f32
- compute_5[(cse_var_2 + 194)] = 0f32
- compute_5[(cse_var_2 + 195)] = 0f32
- compute_5[(cse_var_2 + 196)] = 0f32
- compute_5[(cse_var_2 + 197)] = 0f32
- compute_5[(cse_var_2 + 198)] = 0f32
- compute_5[(cse_var_2 + 199)] = 0f32
- compute_5[(cse_var_2 + 200)] = 0f32
- compute_5[(cse_var_2 + 201)] = 0f32
- compute_5[(cse_var_2 + 202)] = 0f32
- compute_5[(cse_var_2 + 203)] = 0f32
- compute_5[(cse_var_2 + 204)] = 0f32
- compute_5[(cse_var_2 + 205)] = 0f32
- compute_5[(cse_var_2 + 206)] = 0f32
- compute_5[(cse_var_2 + 207)] = 0f32
- compute_5[(cse_var_2 + 224)] = 0f32
- compute_5[(cse_var_2 + 225)] = 0f32
- compute_5[(cse_var_2 + 226)] = 0f32
- compute_5[(cse_var_2 + 227)] = 0f32
- compute_5[(cse_var_2 + 228)] = 0f32
- compute_5[(cse_var_2 + 229)] = 0f32
- compute_5[(cse_var_2 + 230)] = 0f32
- compute_5[(cse_var_2 + 231)] = 0f32
- compute_5[(cse_var_2 + 232)] = 0f32
- compute_5[(cse_var_2 + 233)] = 0f32
- compute_5[(cse_var_2 + 234)] = 0f32
- compute_5[(cse_var_2 + 235)] = 0f32
- compute_5[(cse_var_2 + 236)] = 0f32
- compute_5[(cse_var_2 + 237)] = 0f32
- compute_5[(cse_var_2 + 238)] = 0f32
- compute_5[(cse_var_2 + 239)] = 0f32
- compute_5[(cse_var_2 + 256)] = 0f32
- compute_5[(cse_var_2 + 257)] = 0f32
- compute_5[(cse_var_2 + 258)] = 0f32
- compute_5[(cse_var_2 + 259)] = 0f32
- compute_5[(cse_var_2 + 260)] = 0f32
- compute_5[(cse_var_2 + 261)] = 0f32
- compute_5[(cse_var_2 + 262)] = 0f32
- compute_5[(cse_var_2 + 263)] = 0f32
- compute_5[(cse_var_2 + 264)] = 0f32
- compute_5[(cse_var_2 + 265)] = 0f32
- compute_5[(cse_var_2 + 266)] = 0f32
- compute_5[(cse_var_2 + 267)] = 0f32
- compute_5[(cse_var_2 + 268)] = 0f32
- compute_5[(cse_var_2 + 269)] = 0f32
- compute_5[(cse_var_2 + 270)] = 0f32
- compute_5[(cse_var_2 + 271)] = 0f32
- compute_5[(cse_var_2 + 288)] = 0f32
- compute_5[(cse_var_2 + 289)] = 0f32
- compute_5[(cse_var_2 + 290)] = 0f32
- compute_5[(cse_var_2 + 291)] = 0f32
- compute_5[(cse_var_2 + 292)] = 0f32
- compute_5[(cse_var_2 + 293)] = 0f32
- compute_5[(cse_var_2 + 294)] = 0f32
- compute_5[(cse_var_2 + 295)] = 0f32
- compute_5[(cse_var_2 + 296)] = 0f32
- compute_5[(cse_var_2 + 297)] = 0f32
- compute_5[(cse_var_2 + 298)] = 0f32
- compute_5[(cse_var_2 + 299)] = 0f32
- compute_5[(cse_var_2 + 300)] = 0f32
- compute_5[(cse_var_2 + 301)] = 0f32
- compute_5[(cse_var_2 + 302)] = 0f32
- compute_5[(cse_var_2 + 303)] = 0f32
- compute_5[(cse_var_2 + 320)] = 0f32
- compute_5[(cse_var_2 + 321)] = 0f32
- compute_5[(cse_var_2 + 322)] = 0f32
- compute_5[(cse_var_2 + 323)] = 0f32
- compute_5[(cse_var_2 + 324)] = 0f32
- compute_5[(cse_var_2 + 325)] = 0f32
- compute_5[(cse_var_2 + 326)] = 0f32
- compute_5[(cse_var_2 + 327)] = 0f32
- compute_5[(cse_var_2 + 328)] = 0f32
- compute_5[(cse_var_2 + 329)] = 0f32
- compute_5[(cse_var_2 + 330)] = 0f32
- compute_5[(cse_var_2 + 331)] = 0f32
- compute_5[(cse_var_2 + 332)] = 0f32
- compute_5[(cse_var_2 + 333)] = 0f32
- compute_5[(cse_var_2 + 334)] = 0f32
- compute_5[(cse_var_2 + 335)] = 0f32
- compute_5[(cse_var_2 + 352)] = 0f32
- compute_5[(cse_var_2 + 353)] = 0f32
- compute_5[(cse_var_2 + 354)] = 0f32
- compute_5[(cse_var_2 + 355)] = 0f32
- compute_5[(cse_var_2 + 356)] = 0f32
- compute_5[(cse_var_2 + 357)] = 0f32
- compute_5[(cse_var_2 + 358)] = 0f32
- compute_5[(cse_var_2 + 359)] = 0f32
- compute_5[(cse_var_2 + 360)] = 0f32
- compute_5[(cse_var_2 + 361)] = 0f32
- compute_5[(cse_var_2 + 362)] = 0f32
- compute_5[(cse_var_2 + 363)] = 0f32
- compute_5[(cse_var_2 + 364)] = 0f32
- compute_5[(cse_var_2 + 365)] = 0f32
- compute_5[(cse_var_2 + 366)] = 0f32
- compute_5[(cse_var_2 + 367)] = 0f32
- compute_5[(cse_var_2 + 384)] = 0f32
- compute_5[(cse_var_2 + 385)] = 0f32
- compute_5[(cse_var_2 + 386)] = 0f32
- compute_5[(cse_var_2 + 387)] = 0f32
- compute_5[(cse_var_2 + 388)] = 0f32
- compute_5[(cse_var_2 + 389)] = 0f32
- compute_5[(cse_var_2 + 390)] = 0f32
- compute_5[(cse_var_2 + 391)] = 0f32
- compute_5[(cse_var_2 + 392)] = 0f32
- compute_5[(cse_var_2 + 393)] = 0f32
- compute_5[(cse_var_2 + 394)] = 0f32
- compute_5[(cse_var_2 + 395)] = 0f32
- compute_5[(cse_var_2 + 396)] = 0f32
- compute_5[(cse_var_2 + 397)] = 0f32
- compute_5[(cse_var_2 + 398)] = 0f32
- compute_5[(cse_var_2 + 399)] = 0f32
- compute_5[(cse_var_2 + 416)] = 0f32
- compute_5[(cse_var_2 + 417)] = 0f32
- compute_5[(cse_var_2 + 418)] = 0f32
- compute_5[(cse_var_2 + 419)] = 0f32
- compute_5[(cse_var_2 + 420)] = 0f32
- compute_5[(cse_var_2 + 421)] = 0f32
- compute_5[(cse_var_2 + 422)] = 0f32
- compute_5[(cse_var_2 + 423)] = 0f32
- compute_5[(cse_var_2 + 424)] = 0f32
- compute_5[(cse_var_2 + 425)] = 0f32
- compute_5[(cse_var_2 + 426)] = 0f32
- compute_5[(cse_var_2 + 427)] = 0f32
- compute_5[(cse_var_2 + 428)] = 0f32
- compute_5[(cse_var_2 + 429)] = 0f32
- compute_5[(cse_var_2 + 430)] = 0f32
- compute_5[(cse_var_2 + 431)] = 0f32
- compute_5[(cse_var_2 + 448)] = 0f32
- compute_5[(cse_var_2 + 449)] = 0f32
- compute_5[(cse_var_2 + 450)] = 0f32
- compute_5[(cse_var_2 + 451)] = 0f32
- compute_5[(cse_var_2 + 452)] = 0f32
- compute_5[(cse_var_2 + 453)] = 0f32
- compute_5[(cse_var_2 + 454)] = 0f32
- compute_5[(cse_var_2 + 455)] = 0f32
- compute_5[(cse_var_2 + 456)] = 0f32
- compute_5[(cse_var_2 + 457)] = 0f32
- compute_5[(cse_var_2 + 458)] = 0f32
- compute_5[(cse_var_2 + 459)] = 0f32
- compute_5[(cse_var_2 + 460)] = 0f32
- compute_5[(cse_var_2 + 461)] = 0f32
- compute_5[(cse_var_2 + 462)] = 0f32
- compute_5[(cse_var_2 + 463)] = 0f32
- compute_5[(cse_var_2 + 480)] = 0f32
- compute_5[(cse_var_2 + 481)] = 0f32
- compute_5[(cse_var_2 + 482)] = 0f32
- compute_5[(cse_var_2 + 483)] = 0f32
- compute_5[(cse_var_2 + 484)] = 0f32
- compute_5[(cse_var_2 + 485)] = 0f32
- compute_5[(cse_var_2 + 486)] = 0f32
- compute_5[(cse_var_2 + 487)] = 0f32
- compute_5[(cse_var_2 + 488)] = 0f32
- compute_5[(cse_var_2 + 489)] = 0f32
- compute_5[(cse_var_2 + 490)] = 0f32
- compute_5[(cse_var_2 + 491)] = 0f32
- compute_5[(cse_var_2 + 492)] = 0f32
- compute_5[(cse_var_2 + 493)] = 0f32
- compute_5[(cse_var_2 + 494)] = 0f32
- compute_5[(cse_var_2 + 495)] = 0f32
- for (elem_idx: int32, 0, (placeholder_3[(cse_var_1 + 1)] - placeholder_3[cse_var_1])) {
- let cse_var_259: int32 = (i0.outer*4096)
- let cse_var_258: int32 = (elem_idx*16)
- let cse_var_257: int32 = (cse_var_2 + 99)
- let cse_var_256: int32 = (cse_var_2 + 98)
- let cse_var_255: int32 = (cse_var_2 + 97)
- let cse_var_254: int32 = (cse_var_2 + 96)
- let cse_var_253: int32 = (cse_var_2 + 9)
- let cse_var_252: int32 = (cse_var_2 + 8)
- let cse_var_251: int32 = (cse_var_2 + 79)
- let cse_var_250: int32 = (cse_var_2 + 78)
- let cse_var_249: int32 = (cse_var_2 + 77)
- let cse_var_248: int32 = (cse_var_2 + 76)
- let cse_var_247: int32 = (cse_var_2 + 75)
- let cse_var_246: int32 = (cse_var_2 + 74)
- let cse_var_245: int32 = (cse_var_2 + 73)
- let cse_var_244: int32 = (cse_var_2 + 72)
- let cse_var_243: int32 = (cse_var_2 + 71)
- let cse_var_242: int32 = (cse_var_2 + 70)
- let cse_var_241: int32 = (cse_var_2 + 7)
- let cse_var_240: int32 = (cse_var_2 + 69)
- let cse_var_239: int32 = (cse_var_2 + 68)
- let cse_var_238: int32 = (cse_var_2 + 67)
- let cse_var_237: int32 = (cse_var_2 + 66)
- let cse_var_236: int32 = (cse_var_2 + 65)
- let cse_var_235: int32 = (cse_var_2 + 64)
- let cse_var_234: int32 = (cse_var_2 + 6)
- let cse_var_233: int32 = (cse_var_2 + 5)
- let cse_var_232: int32 = (cse_var_2 + 495)
- let cse_var_231: int32 = (cse_var_2 + 494)
- let cse_var_230: int32 = (cse_var_2 + 493)
- let cse_var_229: int32 = (cse_var_2 + 492)
- let cse_var_228: int32 = (cse_var_2 + 491)
- let cse_var_227: int32 = (cse_var_2 + 490)
- let cse_var_226: int32 = (cse_var_2 + 489)
- let cse_var_225: int32 = (cse_var_2 + 488)
- let cse_var_224: int32 = (cse_var_2 + 487)
- let cse_var_223: int32 = (cse_var_2 + 486)
- let cse_var_222: int32 = (cse_var_2 + 485)
- let cse_var_221: int32 = (cse_var_2 + 484)
- let cse_var_220: int32 = (cse_var_2 + 483)
- let cse_var_219: int32 = (cse_var_2 + 482)
- let cse_var_218: int32 = (cse_var_2 + 481)
- let cse_var_217: int32 = (cse_var_2 + 480)
- let cse_var_216: int32 = (cse_var_2 + 47)
- let cse_var_215: int32 = (cse_var_2 + 463)
- let cse_var_214: int32 = (cse_var_2 + 462)
- let cse_var_213: int32 = (cse_var_2 + 461)
- let cse_var_212: int32 = (cse_var_2 + 460)
- let cse_var_211: int32 = (cse_var_2 + 46)
- let cse_var_210: int32 = (cse_var_2 + 459)
- let cse_var_209: int32 = (cse_var_2 + 458)
- let cse_var_208: int32 = (cse_var_2 + 457)
- let cse_var_207: int32 = (cse_var_2 + 456)
- let cse_var_206: int32 = (cse_var_2 + 455)
- let cse_var_205: int32 = (cse_var_2 + 454)
- let cse_var_204: int32 = (cse_var_2 + 453)
- let cse_var_203: int32 = (cse_var_2 + 452)
- let cse_var_202: int32 = (cse_var_2 + 451)
- let cse_var_201: int32 = (cse_var_2 + 450)
- let cse_var_200: int32 = (cse_var_2 + 45)
- let cse_var_199: int32 = (cse_var_2 + 449)
- let cse_var_198: int32 = (cse_var_2 + 448)
- let cse_var_197: int32 = (cse_var_2 + 44)
- let cse_var_196: int32 = (cse_var_2 + 431)
- let cse_var_195: int32 = (cse_var_2 + 430)
- let cse_var_194: int32 = (cse_var_2 + 43)
- let cse_var_193: int32 = (cse_var_2 + 429)
- let cse_var_192: int32 = (cse_var_2 + 428)
- let cse_var_191: int32 = (cse_var_2 + 427)
- let cse_var_190: int32 = (cse_var_2 + 426)
- let cse_var_189: int32 = (cse_var_2 + 425)
- let cse_var_188: int32 = (cse_var_2 + 424)
- let cse_var_187: int32 = (cse_var_2 + 423)
- let cse_var_186: int32 = (cse_var_2 + 422)
- let cse_var_185: int32 = (cse_var_2 + 421)
- let cse_var_184: int32 = (cse_var_2 + 420)
- let cse_var_183: int32 = (cse_var_2 + 42)
- let cse_var_182: int32 = (cse_var_2 + 419)
- let cse_var_181: int32 = (cse_var_2 + 418)
- let cse_var_180: int32 = (cse_var_2 + 417)
- let cse_var_179: int32 = (cse_var_2 + 416)
- let cse_var_178: int32 = (cse_var_2 + 41)
- let cse_var_177: int32 = (cse_var_2 + 40)
- let cse_var_176: int32 = (cse_var_2 + 4)
- let cse_var_175: int32 = (cse_var_2 + 399)
- let cse_var_174: int32 = (cse_var_2 + 398)
- let cse_var_173: int32 = (cse_var_2 + 397)
- let cse_var_172: int32 = (cse_var_2 + 396)
- let cse_var_171: int32 = (cse_var_2 + 395)
- let cse_var_170: int32 = (cse_var_2 + 394)
- let cse_var_169: int32 = (cse_var_2 + 393)
- let cse_var_168: int32 = (cse_var_2 + 392)
- let cse_var_167: int32 = (cse_var_2 + 391)
- let cse_var_166: int32 = (cse_var_2 + 390)
- let cse_var_165: int32 = (cse_var_2 + 39)
- let cse_var_164: int32 = (cse_var_2 + 389)
- let cse_var_163: int32 = (cse_var_2 + 388)
- let cse_var_162: int32 = (cse_var_2 + 387)
- let cse_var_161: int32 = (cse_var_2 + 386)
- let cse_var_160: int32 = (cse_var_2 + 385)
- let cse_var_159: int32 = (cse_var_2 + 384)
- let cse_var_158: int32 = (cse_var_2 + 38)
- let cse_var_157: int32 = (cse_var_2 + 37)
- let cse_var_156: int32 = (cse_var_2 + 367)
- let cse_var_155: int32 = (cse_var_2 + 366)
- let cse_var_154: int32 = (cse_var_2 + 365)
- let cse_var_153: int32 = (cse_var_2 + 364)
- let cse_var_152: int32 = (cse_var_2 + 363)
- let cse_var_151: int32 = (cse_var_2 + 362)
- let cse_var_150: int32 = (cse_var_2 + 361)
- let cse_var_149: int32 = (cse_var_2 + 360)
- let cse_var_148: int32 = (cse_var_2 + 36)
- let cse_var_147: int32 = (cse_var_2 + 359)
- let cse_var_146: int32 = (cse_var_2 + 358)
- let cse_var_145: int32 = (cse_var_2 + 357)
- let cse_var_144: int32 = (cse_var_2 + 356)
- let cse_var_143: int32 = (cse_var_2 + 355)
- let cse_var_142: int32 = (cse_var_2 + 354)
- let cse_var_141: int32 = (cse_var_2 + 353)
- let cse_var_140: int32 = (cse_var_2 + 352)
- let cse_var_139: int32 = (cse_var_2 + 35)
- let cse_var_138: int32 = (cse_var_2 + 34)
- let cse_var_137: int32 = (cse_var_2 + 335)
- let cse_var_136: int32 = (cse_var_2 + 334)
- let cse_var_135: int32 = (cse_var_2 + 333)
- let cse_var_134: int32 = (cse_var_2 + 332)
- let cse_var_133: int32 = (cse_var_2 + 331)
- let cse_var_132: int32 = (cse_var_2 + 330)
- let cse_var_131: int32 = (cse_var_2 + 33)
- let cse_var_130: int32 = (cse_var_2 + 329)
- let cse_var_129: int32 = (cse_var_2 + 328)
- let cse_var_128: int32 = (cse_var_2 + 327)
- let cse_var_127: int32 = (cse_var_2 + 326)
- let cse_var_126: int32 = (cse_var_2 + 325)
- let cse_var_125: int32 = (cse_var_2 + 324)
- let cse_var_124: int32 = (cse_var_2 + 323)
- let cse_var_123: int32 = (cse_var_2 + 322)
- let cse_var_122: int32 = (cse_var_2 + 321)
- let cse_var_121: int32 = (cse_var_2 + 320)
- let cse_var_120: int32 = (cse_var_2 + 32)
- let cse_var_119: int32 = (cse_var_2 + 303)
- let cse_var_118: int32 = (cse_var_2 + 302)
- let cse_var_117: int32 = (cse_var_2 + 301)
- let cse_var_116: int32 = (cse_var_2 + 300)
- let cse_var_115: int32 = (cse_var_2 + 3)
- let cse_var_114: int32 = (cse_var_2 + 299)
- let cse_var_113: int32 = (cse_var_2 + 298)
- let cse_var_112: int32 = (cse_var_2 + 297)
- let cse_var_111: int32 = (cse_var_2 + 296)
- let cse_var_110: int32 = (cse_var_2 + 295)
- let cse_var_109: int32 = (cse_var_2 + 294)
- let cse_var_108: int32 = (cse_var_2 + 293)
- let cse_var_107: int32 = (cse_var_2 + 292)
- let cse_var_106: int32 = (cse_var_2 + 291)
- let cse_var_105: int32 = (cse_var_2 + 290)
- let cse_var_104: int32 = (cse_var_2 + 289)
- let cse_var_103: int32 = (cse_var_2 + 288)
- let cse_var_102: int32 = (cse_var_2 + 271)
- let cse_var_101: int32 = (cse_var_2 + 270)
- let cse_var_100: int32 = (cse_var_2 + 269)
- let cse_var_99: int32 = (cse_var_2 + 268)
- let cse_var_98: int32 = (cse_var_2 + 267)
- let cse_var_97: int32 = (cse_var_2 + 266)
- let cse_var_96: int32 = (cse_var_2 + 265)
- let cse_var_95: int32 = (cse_var_2 + 264)
- let cse_var_94: int32 = (cse_var_2 + 263)
- let cse_var_93: int32 = (cse_var_2 + 262)
- let cse_var_92: int32 = (cse_var_2 + 261)
- let cse_var_91: int32 = (cse_var_2 + 260)
- let cse_var_90: int32 = (cse_var_2 + 259)
- let cse_var_89: int32 = (cse_var_2 + 258)
- let cse_var_88: int32 = (cse_var_2 + 257)
- let cse_var_87: int32 = (cse_var_2 + 256)
- let cse_var_86: int32 = (cse_var_2 + 239)
- let cse_var_85: int32 = (cse_var_2 + 238)
- let cse_var_84: int32 = (cse_var_2 + 237)
- let cse_var_83: int32 = (cse_var_2 + 236)
- let cse_var_82: int32 = (cse_var_2 + 235)
- let cse_var_81: int32 = (cse_var_2 + 234)
- let cse_var_80: int32 = (cse_var_2 + 233)
- let cse_var_79: int32 = (cse_var_2 + 232)
- let cse_var_78: int32 = (cse_var_2 + 231)
- let cse_var_77: int32 = (cse_var_2 + 230)
- let cse_var_76: int32 = (cse_var_2 + 229)
- let cse_var_75: int32 = (cse_var_2 + 228)
- let cse_var_74: int32 = (cse_var_2 + 227)
- let cse_var_73: int32 = (cse_var_2 + 226)
- let cse_var_72: int32 = (cse_var_2 + 225)
- let cse_var_71: int32 = (cse_var_2 + 224)
- let cse_var_70: int32 = (cse_var_2 + 207)
- let cse_var_69: int32 = (cse_var_2 + 206)
- let cse_var_68: int32 = (cse_var_2 + 205)
- let cse_var_67: int32 = (cse_var_2 + 204)
- let cse_var_66: int32 = (cse_var_2 + 203)
- let cse_var_65: int32 = (cse_var_2 + 202)
- let cse_var_64: int32 = (cse_var_2 + 201)
- let cse_var_63: int32 = (cse_var_2 + 200)
- let cse_var_62: int32 = (cse_var_2 + 2)
- let cse_var_61: int32 = (cse_var_2 + 199)
- let cse_var_60: int32 = (cse_var_2 + 198)
- let cse_var_59: int32 = (cse_var_2 + 197)
- let cse_var_58: int32 = (cse_var_2 + 196)
- let cse_var_57: int32 = (cse_var_2 + 195)
- let cse_var_56: int32 = (cse_var_2 + 194)
- let cse_var_55: int32 = (cse_var_2 + 193)
- let cse_var_54: int32 = (cse_var_2 + 192)
- let cse_var_53: int32 = (cse_var_2 + 175)
- let cse_var_52: int32 = (cse_var_2 + 174)
- let cse_var_51: int32 = (cse_var_2 + 173)
- let cse_var_50: int32 = (cse_var_2 + 172)
- let cse_var_49: int32 = (cse_var_2 + 171)
- let cse_var_48: int32 = (cse_var_2 + 170)
- let cse_var_47: int32 = (cse_var_2 + 169)
- let cse_var_46: int32 = (cse_var_2 + 168)
- let cse_var_45: int32 = (cse_var_2 + 167)
- let cse_var_44: int32 = (cse_var_2 + 166)
- let cse_var_43: int32 = (cse_var_2 + 165)
- let cse_var_42: int32 = (cse_var_2 + 164)
- let cse_var_41: int32 = (cse_var_2 + 163)
- let cse_var_40: int32 = (cse_var_2 + 162)
- let cse_var_39: int32 = (cse_var_2 + 161)
- let cse_var_38: int32 = (cse_var_2 + 160)
- let cse_var_37: int32 = (cse_var_2 + 15)
- let cse_var_36: int32 = (cse_var_2 + 143)
- let cse_var_35: int32 = (cse_var_2 + 142)
- let cse_var_34: int32 = (cse_var_2 + 141)
- let cse_var_33: int32 = (cse_var_2 + 140)
- let cse_var_32: int32 = (cse_var_2 + 14)
- let cse_var_31: int32 = (cse_var_2 + 139)
- let cse_var_30: int32 = (cse_var_2 + 138)
- let cse_var_29: int32 = (cse_var_2 + 137)
- let cse_var_28: int32 = (cse_var_2 + 136)
- let cse_var_27: int32 = (cse_var_2 + 135)
- let cse_var_26: int32 = (cse_var_2 + 134)
- let cse_var_25: int32 = (cse_var_2 + 133)
- let cse_var_24: int32 = (cse_var_2 + 132)
- let cse_var_23: int32 = (cse_var_2 + 131)
- let cse_var_22: int32 = (cse_var_2 + 130)
- let cse_var_21: int32 = (cse_var_2 + 13)
- let cse_var_20: int32 = (cse_var_2 + 129)
- let cse_var_19: int32 = (cse_var_2 + 128)
- let cse_var_18: int32 = (cse_var_2 + 12)
- let cse_var_17: int32 = (cse_var_2 + 111)
- let cse_var_16: int32 = (cse_var_2 + 110)
- let cse_var_15: int32 = (cse_var_2 + 11)
- let cse_var_14: int32 = (cse_var_2 + 109)
- let cse_var_13: int32 = (cse_var_2 + 108)
- let cse_var_12: int32 = (cse_var_2 + 107)
- let cse_var_11: int32 = (cse_var_2 + 106)
- let cse_var_10: int32 = (cse_var_2 + 105)
- let cse_var_9: int32 = (cse_var_2 + 104)
- let cse_var_8: int32 = (cse_var_2 + 103)
- let cse_var_7: int32 = (cse_var_2 + 102)
- let cse_var_6: int32 = (cse_var_2 + 101)
- let cse_var_5: int32 = (cse_var_2 + 100)
- let cse_var_4: int32 = (cse_var_2 + 10)
- let cse_var_3: int32 = (cse_var_2 + 1)
- {
- compute_5[cse_var_2] = (compute_5[cse_var_2] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_3] = (compute_5[cse_var_3] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_62] = (compute_5[cse_var_62] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_115] = (compute_5[cse_var_115] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_176] = (compute_5[cse_var_176] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_233] = (compute_5[cse_var_233] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_234] = (compute_5[cse_var_234] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_241] = (compute_5[cse_var_241] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_252] = (compute_5[cse_var_252] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_253] = (compute_5[cse_var_253] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_4] = (compute_5[cse_var_4] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_21] = (compute_5[cse_var_21] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_32] = (compute_5[cse_var_32] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_37] = (compute_5[cse_var_37] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[(cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)])], 0f32)))
- compute_5[cse_var_120] = (compute_5[cse_var_120] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_131] = (compute_5[cse_var_131] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_138] = (compute_5[cse_var_138] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_139] = (compute_5[cse_var_139] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_148] = (compute_5[cse_var_148] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_157] = (compute_5[cse_var_157] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_158] = (compute_5[cse_var_158] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_165] = (compute_5[cse_var_165] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_177] = (compute_5[cse_var_177] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_178] = (compute_5[cse_var_178] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_183] = (compute_5[cse_var_183] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_194] = (compute_5[cse_var_194] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_197] = (compute_5[cse_var_197] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_200] = (compute_5[cse_var_200] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_211] = (compute_5[cse_var_211] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_216] = (compute_5[cse_var_216] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 256)], 0f32)))
- compute_5[cse_var_235] = (compute_5[cse_var_235] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_236] = (compute_5[cse_var_236] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_237] = (compute_5[cse_var_237] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_238] = (compute_5[cse_var_238] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_239] = (compute_5[cse_var_239] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_240] = (compute_5[cse_var_240] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_242] = (compute_5[cse_var_242] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_243] = (compute_5[cse_var_243] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_244] = (compute_5[cse_var_244] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_245] = (compute_5[cse_var_245] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_246] = (compute_5[cse_var_246] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_247] = (compute_5[cse_var_247] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_248] = (compute_5[cse_var_248] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_249] = (compute_5[cse_var_249] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_250] = (compute_5[cse_var_250] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_251] = (compute_5[cse_var_251] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 512)], 0f32)))
- compute_5[cse_var_254] = (compute_5[cse_var_254] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_255] = (compute_5[cse_var_255] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_256] = (compute_5[cse_var_256] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_257] = (compute_5[cse_var_257] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 768)], 0f32)))
- compute_5[cse_var_19] = (compute_5[cse_var_19] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_20] = (compute_5[cse_var_20] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_22] = (compute_5[cse_var_22] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_23] = (compute_5[cse_var_23] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_24] = (compute_5[cse_var_24] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_25] = (compute_5[cse_var_25] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_26] = (compute_5[cse_var_26] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_27] = (compute_5[cse_var_27] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_28] = (compute_5[cse_var_28] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_29] = (compute_5[cse_var_29] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_30] = (compute_5[cse_var_30] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_31] = (compute_5[cse_var_31] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_33] = (compute_5[cse_var_33] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_34] = (compute_5[cse_var_34] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_35] = (compute_5[cse_var_35] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_36] = (compute_5[cse_var_36] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1024)], 0f32)))
- compute_5[cse_var_38] = (compute_5[cse_var_38] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_39] = (compute_5[cse_var_39] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_40] = (compute_5[cse_var_40] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_41] = (compute_5[cse_var_41] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_42] = (compute_5[cse_var_42] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_43] = (compute_5[cse_var_43] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_44] = (compute_5[cse_var_44] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_45] = (compute_5[cse_var_45] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_46] = (compute_5[cse_var_46] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_47] = (compute_5[cse_var_47] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_48] = (compute_5[cse_var_48] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_49] = (compute_5[cse_var_49] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_50] = (compute_5[cse_var_50] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_51] = (compute_5[cse_var_51] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_52] = (compute_5[cse_var_52] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_53] = (compute_5[cse_var_53] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1280)], 0f32)))
- compute_5[cse_var_54] = (compute_5[cse_var_54] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_55] = (compute_5[cse_var_55] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_56] = (compute_5[cse_var_56] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_57] = (compute_5[cse_var_57] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_58] = (compute_5[cse_var_58] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_59] = (compute_5[cse_var_59] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_60] = (compute_5[cse_var_60] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_61] = (compute_5[cse_var_61] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_63] = (compute_5[cse_var_63] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_64] = (compute_5[cse_var_64] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_65] = (compute_5[cse_var_65] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_66] = (compute_5[cse_var_66] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_67] = (compute_5[cse_var_67] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_68] = (compute_5[cse_var_68] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_69] = (compute_5[cse_var_69] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_70] = (compute_5[cse_var_70] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1536)], 0f32)))
- compute_5[cse_var_71] = (compute_5[cse_var_71] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_72] = (compute_5[cse_var_72] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_73] = (compute_5[cse_var_73] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_74] = (compute_5[cse_var_74] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_75] = (compute_5[cse_var_75] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_76] = (compute_5[cse_var_76] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_77] = (compute_5[cse_var_77] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_78] = (compute_5[cse_var_78] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_79] = (compute_5[cse_var_79] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_80] = (compute_5[cse_var_80] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_81] = (compute_5[cse_var_81] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_82] = (compute_5[cse_var_82] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_83] = (compute_5[cse_var_83] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_84] = (compute_5[cse_var_84] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_85] = (compute_5[cse_var_85] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_86] = (compute_5[cse_var_86] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 1792)], 0f32)))
- compute_5[cse_var_87] = (compute_5[cse_var_87] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_88] = (compute_5[cse_var_88] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_89] = (compute_5[cse_var_89] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_90] = (compute_5[cse_var_90] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_91] = (compute_5[cse_var_91] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_92] = (compute_5[cse_var_92] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_93] = (compute_5[cse_var_93] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_94] = (compute_5[cse_var_94] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_95] = (compute_5[cse_var_95] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_96] = (compute_5[cse_var_96] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_97] = (compute_5[cse_var_97] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_98] = (compute_5[cse_var_98] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_99] = (compute_5[cse_var_99] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_100] = (compute_5[cse_var_100] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_101] = (compute_5[cse_var_101] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_102] = (compute_5[cse_var_102] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2048)], 0f32)))
- compute_5[cse_var_103] = (compute_5[cse_var_103] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_104] = (compute_5[cse_var_104] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_105] = (compute_5[cse_var_105] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_106] = (compute_5[cse_var_106] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_107] = (compute_5[cse_var_107] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_108] = (compute_5[cse_var_108] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_109] = (compute_5[cse_var_109] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_110] = (compute_5[cse_var_110] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_111] = (compute_5[cse_var_111] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_112] = (compute_5[cse_var_112] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_113] = (compute_5[cse_var_113] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_114] = (compute_5[cse_var_114] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_116] = (compute_5[cse_var_116] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_117] = (compute_5[cse_var_117] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_118] = (compute_5[cse_var_118] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_119] = (compute_5[cse_var_119] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2304)], 0f32)))
- compute_5[cse_var_121] = (compute_5[cse_var_121] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_122] = (compute_5[cse_var_122] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_123] = (compute_5[cse_var_123] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_124] = (compute_5[cse_var_124] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_125] = (compute_5[cse_var_125] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_126] = (compute_5[cse_var_126] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_127] = (compute_5[cse_var_127] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_128] = (compute_5[cse_var_128] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_129] = (compute_5[cse_var_129] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_130] = (compute_5[cse_var_130] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_132] = (compute_5[cse_var_132] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_133] = (compute_5[cse_var_133] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_134] = (compute_5[cse_var_134] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_135] = (compute_5[cse_var_135] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_136] = (compute_5[cse_var_136] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_137] = (compute_5[cse_var_137] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2560)], 0f32)))
- compute_5[cse_var_140] = (compute_5[cse_var_140] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_141] = (compute_5[cse_var_141] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_142] = (compute_5[cse_var_142] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_143] = (compute_5[cse_var_143] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_144] = (compute_5[cse_var_144] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_145] = (compute_5[cse_var_145] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_146] = (compute_5[cse_var_146] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_147] = (compute_5[cse_var_147] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_149] = (compute_5[cse_var_149] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_150] = (compute_5[cse_var_150] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_151] = (compute_5[cse_var_151] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_152] = (compute_5[cse_var_152] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_153] = (compute_5[cse_var_153] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_154] = (compute_5[cse_var_154] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_155] = (compute_5[cse_var_155] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_156] = (compute_5[cse_var_156] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 2816)], 0f32)))
- compute_5[cse_var_159] = (compute_5[cse_var_159] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_160] = (compute_5[cse_var_160] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_161] = (compute_5[cse_var_161] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_162] = (compute_5[cse_var_162] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_163] = (compute_5[cse_var_163] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_164] = (compute_5[cse_var_164] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_166] = (compute_5[cse_var_166] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_167] = (compute_5[cse_var_167] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_168] = (compute_5[cse_var_168] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_169] = (compute_5[cse_var_169] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_170] = (compute_5[cse_var_170] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_171] = (compute_5[cse_var_171] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_172] = (compute_5[cse_var_172] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_173] = (compute_5[cse_var_173] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_174] = (compute_5[cse_var_174] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_175] = (compute_5[cse_var_175] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3072)], 0f32)))
- compute_5[cse_var_179] = (compute_5[cse_var_179] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_180] = (compute_5[cse_var_180] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_181] = (compute_5[cse_var_181] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_182] = (compute_5[cse_var_182] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_184] = (compute_5[cse_var_184] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_185] = (compute_5[cse_var_185] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_186] = (compute_5[cse_var_186] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_187] = (compute_5[cse_var_187] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_188] = (compute_5[cse_var_188] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_189] = (compute_5[cse_var_189] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_190] = (compute_5[cse_var_190] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_191] = (compute_5[cse_var_191] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_192] = (compute_5[cse_var_192] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_193] = (compute_5[cse_var_193] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_195] = (compute_5[cse_var_195] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_196] = (compute_5[cse_var_196] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3328)], 0f32)))
- compute_5[cse_var_198] = (compute_5[cse_var_198] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_199] = (compute_5[cse_var_199] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_201] = (compute_5[cse_var_201] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_202] = (compute_5[cse_var_202] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_203] = (compute_5[cse_var_203] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_204] = (compute_5[cse_var_204] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_205] = (compute_5[cse_var_205] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_206] = (compute_5[cse_var_206] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_207] = (compute_5[cse_var_207] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_208] = (compute_5[cse_var_208] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_209] = (compute_5[cse_var_209] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_210] = (compute_5[cse_var_210] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_212] = (compute_5[cse_var_212] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_213] = (compute_5[cse_var_213] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_214] = (compute_5[cse_var_214] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_215] = (compute_5[cse_var_215] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3584)], 0f32)))
- compute_5[cse_var_217] = (compute_5[cse_var_217] + (placeholder_1[((placeholder_3[cse_var_1]*16) + cse_var_258)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_218] = (compute_5[cse_var_218] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 1)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_219] = (compute_5[cse_var_219] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 2)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_220] = (compute_5[cse_var_220] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 3)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_221] = (compute_5[cse_var_221] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 4)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_222] = (compute_5[cse_var_222] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 5)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_223] = (compute_5[cse_var_223] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 6)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_224] = (compute_5[cse_var_224] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 7)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_225] = (compute_5[cse_var_225] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 8)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_226] = (compute_5[cse_var_226] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 9)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_227] = (compute_5[cse_var_227] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 10)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_228] = (compute_5[cse_var_228] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 11)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_229] = (compute_5[cse_var_229] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 12)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_230] = (compute_5[cse_var_230] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 13)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_231] = (compute_5[cse_var_231] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 14)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
- compute_5[cse_var_232] = (compute_5[cse_var_232] + (placeholder_1[(((placeholder_3[cse_var_1]*16) + cse_var_258) + 15)]*max(placeholder[((cse_var_259 + placeholder_2[(placeholder_3[cse_var_1] + elem_idx)]) + 3840)], 0f32)))
+ preflattened_buffer_map = {placeholder_6: placeholder_15: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_8: placeholder_16: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_17: Buffer(placeholder_12, int32, [4916], []), placeholder_9: placeholder_18: Buffer(placeholder_14, float32, [128, 512], []), placeholder_5: placeholder_19: Buffer(placeholder_10, float32, [128, 256], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], [])} {
+ for (i0.outer.i1.outer.fused: int32, 0, 16) "parallel" {
+ allocate(compute_4: Pointer(global float32), float32, [4096]), storage_scope = global {
+ for (i.outer.inner: int32, 0, 16) {
+ for (nb_j.inner: int32, 0, 2) {
+ let cse_var_2: int32 = ((i0.outer.i1.outer.fused*2) + nb_j.inner)
+ let cse_var_1: int32 = ((i.outer.inner*256) + (nb_j.inner*16))
+ {
+ compute_5: Buffer(compute_4, float32, [4096], [])[cse_var_1] = 0f32
+ compute_5[(cse_var_1 + 1)] = 0f32
+ compute_5[(cse_var_1 + 2)] = 0f32
+ compute_5[(cse_var_1 + 3)] = 0f32
+ compute_5[(cse_var_1 + 4)] = 0f32
+ compute_5[(cse_var_1 + 5)] = 0f32
+ compute_5[(cse_var_1 + 6)] = 0f32
+ compute_5[(cse_var_1 + 7)] = 0f32
+ compute_5[(cse_var_1 + 8)] = 0f32
+ compute_5[(cse_var_1 + 9)] = 0f32
+ compute_5[(cse_var_1 + 10)] = 0f32
+ compute_5[(cse_var_1 + 11)] = 0f32
+ compute_5[(cse_var_1 + 12)] = 0f32
+ compute_5[(cse_var_1 + 13)] = 0f32
+ compute_5[(cse_var_1 + 14)] = 0f32
+ compute_5[(cse_var_1 + 15)] = 0f32
+ compute_5[(cse_var_1 + 32)] = 0f32
+ compute_5[(cse_var_1 + 33)] = 0f32
+ compute_5[(cse_var_1 + 34)] = 0f32
+ compute_5[(cse_var_1 + 35)] = 0f32
+ compute_5[(cse_var_1 + 36)] = 0f32
+ compute_5[(cse_var_1 + 37)] = 0f32
+ compute_5[(cse_var_1 + 38)] = 0f32
+ compute_5[(cse_var_1 + 39)] = 0f32
+ compute_5[(cse_var_1 + 40)] = 0f32
+ compute_5[(cse_var_1 + 41)] = 0f32
+ compute_5[(cse_var_1 + 42)] = 0f32
+ compute_5[(cse_var_1 + 43)] = 0f32
+ compute_5[(cse_var_1 + 44)] = 0f32
+ compute_5[(cse_var_1 + 45)] = 0f32
+ compute_5[(cse_var_1 + 46)] = 0f32
+ compute_5[(cse_var_1 + 47)] = 0f32
+ compute_5[(cse_var_1 + 64)] = 0f32
+ compute_5[(cse_var_1 + 65)] = 0f32
+ compute_5[(cse_var_1 + 66)] = 0f32
+ compute_5[(cse_var_1 + 67)] = 0f32
+ compute_5[(cse_var_1 + 68)] = 0f32
+ compute_5[(cse_var_1 + 69)] = 0f32
+ compute_5[(cse_var_1 + 70)] = 0f32
+ compute_5[(cse_var_1 + 71)] = 0f32
+ compute_5[(cse_var_1 + 72)] = 0f32
+ compute_5[(cse_var_1 + 73)] = 0f32
+ compute_5[(cse_var_1 + 74)] = 0f32
+ compute_5[(cse_var_1 + 75)] = 0f32
+ compute_5[(cse_var_1 + 76)] = 0f32
+ compute_5[(cse_var_1 + 77)] = 0f32
+ compute_5[(cse_var_1 + 78)] = 0f32
+ compute_5[(cse_var_1 + 79)] = 0f32
+ compute_5[(cse_var_1 + 96)] = 0f32
+ compute_5[(cse_var_1 + 97)] = 0f32
+ compute_5[(cse_var_1 + 98)] = 0f32
+ compute_5[(cse_var_1 + 99)] = 0f32
+ compute_5[(cse_var_1 + 100)] = 0f32
+ compute_5[(cse_var_1 + 101)] = 0f32
+ compute_5[(cse_var_1 + 102)] = 0f32
+ compute_5[(cse_var_1 + 103)] = 0f32
+ compute_5[(cse_var_1 + 104)] = 0f32
+ compute_5[(cse_var_1 + 105)] = 0f32
+ compute_5[(cse_var_1 + 106)] = 0f32
+ compute_5[(cse_var_1 + 107)] = 0f32
+ compute_5[(cse_var_1 + 108)] = 0f32
+ compute_5[(cse_var_1 + 109)] = 0f32
+ compute_5[(cse_var_1 + 110)] = 0f32
+ compute_5[(cse_var_1 + 111)] = 0f32
+ compute_5[(cse_var_1 + 128)] = 0f32
+ compute_5[(cse_var_1 + 129)] = 0f32
+ compute_5[(cse_var_1 + 130)] = 0f32
+ compute_5[(cse_var_1 + 131)] = 0f32
+ compute_5[(cse_var_1 + 132)] = 0f32
+ compute_5[(cse_var_1 + 133)] = 0f32
+ compute_5[(cse_var_1 + 134)] = 0f32
+ compute_5[(cse_var_1 + 135)] = 0f32
+ compute_5[(cse_var_1 + 136)] = 0f32
+ compute_5[(cse_var_1 + 137)] = 0f32
+ compute_5[(cse_var_1 + 138)] = 0f32
+ compute_5[(cse_var_1 + 139)] = 0f32
+ compute_5[(cse_var_1 + 140)] = 0f32
+ compute_5[(cse_var_1 + 141)] = 0f32
+ compute_5[(cse_var_1 + 142)] = 0f32
+ compute_5[(cse_var_1 + 143)] = 0f32
+ compute_5[(cse_var_1 + 160)] = 0f32
+ compute_5[(cse_var_1 + 161)] = 0f32
+ compute_5[(cse_var_1 + 162)] = 0f32
+ compute_5[(cse_var_1 + 163)] = 0f32
+ compute_5[(cse_var_1 + 164)] = 0f32
+ compute_5[(cse_var_1 + 165)] = 0f32
+ compute_5[(cse_var_1 + 166)] = 0f32
+ compute_5[(cse_var_1 + 167)] = 0f32
+ compute_5[(cse_var_1 + 168)] = 0f32
+ compute_5[(cse_var_1 + 169)] = 0f32
+ compute_5[(cse_var_1 + 170)] = 0f32
+ compute_5[(cse_var_1 + 171)] = 0f32
+ compute_5[(cse_var_1 + 172)] = 0f32
+ compute_5[(cse_var_1 + 173)] = 0f32
+ compute_5[(cse_var_1 + 174)] = 0f32
+ compute_5[(cse_var_1 + 175)] = 0f32
+ compute_5[(cse_var_1 + 192)] = 0f32
+ compute_5[(cse_var_1 + 193)] = 0f32
+ compute_5[(cse_var_1 + 194)] = 0f32
+ compute_5[(cse_var_1 + 195)] = 0f32
+ compute_5[(cse_var_1 + 196)] = 0f32
+ compute_5[(cse_var_1 + 197)] = 0f32
+ compute_5[(cse_var_1 + 198)] = 0f32
+ compute_5[(cse_var_1 + 199)] = 0f32
+ compute_5[(cse_var_1 + 200)] = 0f32
+ compute_5[(cse_var_1 + 201)] = 0f32
+ compute_5[(cse_var_1 + 202)] = 0f32
+ compute_5[(cse_var_1 + 203)] = 0f32
+ compute_5[(cse_var_1 + 204)] = 0f32
+ compute_5[(cse_var_1 + 205)] = 0f32
+ compute_5[(cse_var_1 + 206)] = 0f32
+ compute_5[(cse_var_1 + 207)] = 0f32
+ compute_5[(cse_var_1 + 224)] = 0f32
+ compute_5[(cse_var_1 + 225)] = 0f32
+ compute_5[(cse_var_1 + 226)] = 0f32
+ compute_5[(cse_var_1 + 227)] = 0f32
+ compute_5[(cse_var_1 + 228)] = 0f32
+ compute_5[(cse_var_1 + 229)] = 0f32
+ compute_5[(cse_var_1 + 230)] = 0f32
+ compute_5[(cse_var_1 + 231)] = 0f32
+ compute_5[(cse_var_1 + 232)] = 0f32
+ compute_5[(cse_var_1 + 233)] = 0f32
+ compute_5[(cse_var_1 + 234)] = 0f32
+ compute_5[(cse_var_1 + 235)] = 0f32
+ compute_5[(cse_var_1 + 236)] = 0f32
+ compute_5[(cse_var_1 + 237)] = 0f32
+ compute_5[(cse_var_1 + 238)] = 0f32
+ compute_5[(cse_var_1 + 239)] = 0f32
+ for (elem_idx: int32, 0, (placeholder_3[(cse_var_2 + 1)] - placeholder_3[cse_var_2])) {
+ let cse_var_131: int32 = (i.outer.inner*2048)
+ let cse_var_130: int32 = (elem_idx*16)
+ let cse_var_129: int32 = (cse_var_1 + 99)
+ let cse_var_128: int32 = (cse_var_1 + 98)
+ let cse_var_127: int32 = (cse_var_1 + 97)
+ let cse_var_126: int32 = (cse_var_1 + 96)
+ let cse_var_125: int32 = (cse_var_1 + 9)
+ let cse_var_124: int32 = (cse_var_1 + 8)
+ let cse_var_123: int32 = (cse_var_1 + 79)
+ let cse_var_122: int32 = (cse_var_1 + 78)
+ let cse_var_121: int32 = (cse_var_1 + 77)
+ let cse_var_120: int32 = (cse_var_1 + 76)
+ let cse_var_119: int32 = (cse_var_1 + 75)
+ let cse_var_118: int32 = (cse_var_1 + 74)
+ let cse_var_117: int32 = (cse_var_1 + 73)
+ let cse_var_116: int32 = (cse_var_1 + 72)
+ let cse_var_115: int32 = (cse_var_1 + 71)
+ let cse_var_114: int32 = (cse_var_1 + 70)
+ let cse_var_113: int32 = (cse_var_1 + 7)
+ let cse_var_112: int32 = (cse_var_1 + 69)
+ let cse_var_111: int32 = (cse_var_1 + 68)
+ let cse_var_110: int32 = (cse_var_1 + 67)
+ let cse_var_109: int32 = (cse_var_1 + 66)
+ let cse_var_108: int32 = (cse_var_1 + 65)
+ let cse_var_107: int32 = (cse_var_1 + 64)
+ let cse_var_106: int32 = (cse_var_1 + 6)
+ let cse_var_105: int32 = (cse_var_1 + 5)
+ let cse_var_104: int32 = (cse_var_1 + 47)
+ let cse_var_103: int32 = (cse_var_1 + 46)
+ let cse_var_102: int32 = (cse_var_1 + 45)
+ let cse_var_101: int32 = (cse_var_1 + 44)
+ let cse_var_100: int32 = (cse_var_1 + 43)
+ let cse_var_99: int32 = (cse_var_1 + 42)
+ let cse_var_98: int32 = (cse_var_1 + 41)
+ let cse_var_97: int32 = (cse_var_1 + 40)
+ let cse_var_96: int32 = (cse_var_1 + 4)
+ let cse_var_95: int32 = (cse_var_1 + 39)
+ let cse_var_94: int32 = (cse_var_1 + 38)
+ let cse_var_93: int32 = (cse_var_1 + 37)
+ let cse_var_92: int32 = (cse_var_1 + 36)
+ let cse_var_91: int32 = (cse_var_1 + 35)
+ let cse_var_90: int32 = (cse_var_1 + 34)
+ let cse_var_89: int32 = (cse_var_1 + 33)
+ let cse_var_88: int32 = (cse_var_1 + 32)
+ let cse_var_87: int32 = (cse_var_1 + 3)
+ let cse_var_86: int32 = (cse_var_1 + 239)
+ let cse_var_85: int32 = (cse_var_1 + 238)
+ let cse_var_84: int32 = (cse_var_1 + 237)
+ let cse_var_83: int32 = (cse_var_1 + 236)
+ let cse_var_82: int32 = (cse_var_1 + 235)
+ let cse_var_81: int32 = (cse_var_1 + 234)
+ let cse_var_80: int32 = (cse_var_1 + 233)
+ let cse_var_79: int32 = (cse_var_1 + 232)
+ let cse_var_78: int32 = (cse_var_1 + 231)
+ let cse_var_77: int32 = (cse_var_1 + 230)
+ let cse_var_76: int32 = (cse_var_1 + 229)
+ let cse_var_75: int32 = (cse_var_1 + 228)
+ let cse_var_74: int32 = (cse_var_1 + 227)
+ let cse_var_73: int32 = (cse_var_1 + 226)
+ let cse_var_72: int32 = (cse_var_1 + 225)
+ let cse_var_71: int32 = (cse_var_1 + 224)
+ let cse_var_70: int32 = (cse_var_1 + 207)
+ let cse_var_69: int32 = (cse_var_1 + 206)
+ let cse_var_68: int32 = (cse_var_1 + 205)
+ let cse_var_67: int32 = (cse_var_1 + 204)
+ let cse_var_66: int32 = (cse_var_1 + 203)
+ let cse_var_65: int32 = (cse_var_1 + 202)
+ let cse_var_64: int32 = (cse_var_1 + 201)
+ let cse_var_63: int32 = (cse_var_1 + 200)
+ let cse_var_62: int32 = (cse_var_1 + 2)
+ let cse_var_61: int32 = (cse_var_1 + 199)
+ let cse_var_60: int32 = (cse_var_1 + 198)
+ let cse_var_59: int32 = (cse_var_1 + 197)
+ let cse_var_58: int32 = (cse_var_1 + 196)
+ let cse_var_57: int32 = (cse_var_1 + 195)
+ let cse_var_56: int32 = (cse_var_1 + 194)
+ let cse_var_55: int32 = (cse_var_1 + 193)
+ let cse_var_54: int32 = (cse_var_1 + 192)
+ let cse_var_53: int32 = (cse_var_1 + 175)
+ let cse_var_52: int32 = (cse_var_1 + 174)
+ let cse_var_51: int32 = (cse_var_1 + 173)
+ let cse_var_50: int32 = (cse_var_1 + 172)
+ let cse_var_49: int32 = (cse_var_1 + 171)
+ let cse_var_48: int32 = (cse_var_1 + 170)
+ let cse_var_47: int32 = (cse_var_1 + 169)
+ let cse_var_46: int32 = (cse_var_1 + 168)
+ let cse_var_45: int32 = (cse_var_1 + 167)
+ let cse_var_44: int32 = (cse_var_1 + 166)
+ let cse_var_43: int32 = (cse_var_1 + 165)
+ let cse_var_42: int32 = (cse_var_1 + 164)
+ let cse_var_41: int32 = (cse_var_1 + 163)
+ let cse_var_40: int32 = (cse_var_1 + 162)
+ let cse_var_39: int32 = (cse_var_1 + 161)
+ let cse_var_38: int32 = (cse_var_1 + 160)
+ let cse_var_37: int32 = (cse_var_1 + 15)
+ let cse_var_36: int32 = (cse_var_1 + 143)
+ let cse_var_35: int32 = (cse_var_1 + 142)
+ let cse_var_34: int32 = (cse_var_1 + 141)
+ let cse_var_33: int32 = (cse_var_1 + 140)
+ let cse_var_32: int32 = (cse_var_1 + 14)
+ let cse_var_31: int32 = (cse_var_1 + 139)
+ let cse_var_30: int32 = (cse_var_1 + 138)
+ let cse_var_29: int32 = (cse_var_1 + 137)
+ let cse_var_28: int32 = (cse_var_1 + 136)
+ let cse_var_27: int32 = (cse_var_1 + 135)
+ let cse_var_26: int32 = (cse_var_1 + 134)
+ let cse_var_25: int32 = (cse_var_1 + 133)
+ let cse_var_24: int32 = (cse_var_1 + 132)
+ let cse_var_23: int32 = (cse_var_1 + 131)
+ let cse_var_22: int32 = (cse_var_1 + 130)
+ let cse_var_21: int32 = (cse_var_1 + 13)
+ let cse_var_20: int32 = (cse_var_1 + 129)
+ let cse_var_19: int32 = (cse_var_1 + 128)
+ let cse_var_18: int32 = (cse_var_1 + 12)
+ let cse_var_17: int32 = (cse_var_1 + 111)
+ let cse_var_16: int32 = (cse_var_1 + 110)
+ let cse_var_15: int32 = (cse_var_1 + 11)
+ let cse_var_14: int32 = (cse_var_1 + 109)
+ let cse_var_13: int32 = (cse_var_1 + 108)
+ let cse_var_12: int32 = (cse_var_1 + 107)
+ let cse_var_11: int32 = (cse_var_1 + 106)
+ let cse_var_10: int32 = (cse_var_1 + 105)
+ let cse_var_9: int32 = (cse_var_1 + 104)
+ let cse_var_8: int32 = (cse_var_1 + 103)
+ let cse_var_7: int32 = (cse_var_1 + 102)
+ let cse_var_6: int32 = (cse_var_1 + 101)
+ let cse_var_5: int32 = (cse_var_1 + 100)
+ let cse_var_4: int32 = (cse_var_1 + 10)
+ let cse_var_3: int32 = (cse_var_1 + 1)
+ {
+ compute_5[cse_var_1] = (compute_5[cse_var_1] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_3] = (compute_5[cse_var_3] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_62] = (compute_5[cse_var_62] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_87] = (compute_5[cse_var_87] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_96] = (compute_5[cse_var_96] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_105] = (compute_5[cse_var_105] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_106] = (compute_5[cse_var_106] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_113] = (compute_5[cse_var_113] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_124] = (compute_5[cse_var_124] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_125] = (compute_5[cse_var_125] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_4] = (compute_5[cse_var_4] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_21] = (compute_5[cse_var_21] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_32] = (compute_5[cse_var_32] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_37] = (compute_5[cse_var_37] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[(cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)])], 0f32)))
+ compute_5[cse_var_88] = (compute_5[cse_var_88] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_89] = (compute_5[cse_var_89] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_90] = (compute_5[cse_var_90] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_91] = (compute_5[cse_var_91] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_92] = (compute_5[cse_var_92] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_93] = (compute_5[cse_var_93] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_94] = (compute_5[cse_var_94] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_95] = (compute_5[cse_var_95] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_97] = (compute_5[cse_var_97] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_98] = (compute_5[cse_var_98] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_99] = (compute_5[cse_var_99] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_100] = (compute_5[cse_var_100] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_101] = (compute_5[cse_var_101] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_102] = (compute_5[cse_var_102] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_103] = (compute_5[cse_var_103] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_104] = (compute_5[cse_var_104] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 256)], 0f32)))
+ compute_5[cse_var_107] = (compute_5[cse_var_107] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_108] = (compute_5[cse_var_108] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_109] = (compute_5[cse_var_109] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_110] = (compute_5[cse_var_110] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_111] = (compute_5[cse_var_111] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_112] = (compute_5[cse_var_112] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_114] = (compute_5[cse_var_114] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_115] = (compute_5[cse_var_115] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_116] = (compute_5[cse_var_116] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_117] = (compute_5[cse_var_117] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_118] = (compute_5[cse_var_118] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_119] = (compute_5[cse_var_119] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_120] = (compute_5[cse_var_120] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_121] = (compute_5[cse_var_121] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_122] = (compute_5[cse_var_122] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_123] = (compute_5[cse_var_123] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 512)], 0f32)))
+ compute_5[cse_var_126] = (compute_5[cse_var_126] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_127] = (compute_5[cse_var_127] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_128] = (compute_5[cse_var_128] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_129] = (compute_5[cse_var_129] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 768)], 0f32)))
+ compute_5[cse_var_19] = (compute_5[cse_var_19] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_20] = (compute_5[cse_var_20] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_22] = (compute_5[cse_var_22] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_23] = (compute_5[cse_var_23] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_24] = (compute_5[cse_var_24] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_25] = (compute_5[cse_var_25] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_26] = (compute_5[cse_var_26] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_27] = (compute_5[cse_var_27] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_28] = (compute_5[cse_var_28] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_29] = (compute_5[cse_var_29] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_30] = (compute_5[cse_var_30] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_31] = (compute_5[cse_var_31] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_33] = (compute_5[cse_var_33] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_34] = (compute_5[cse_var_34] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_35] = (compute_5[cse_var_35] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_36] = (compute_5[cse_var_36] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1024)], 0f32)))
+ compute_5[cse_var_38] = (compute_5[cse_var_38] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_39] = (compute_5[cse_var_39] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_40] = (compute_5[cse_var_40] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_41] = (compute_5[cse_var_41] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_42] = (compute_5[cse_var_42] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_43] = (compute_5[cse_var_43] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_44] = (compute_5[cse_var_44] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_45] = (compute_5[cse_var_45] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_46] = (compute_5[cse_var_46] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_47] = (compute_5[cse_var_47] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_48] = (compute_5[cse_var_48] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_49] = (compute_5[cse_var_49] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_50] = (compute_5[cse_var_50] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_51] = (compute_5[cse_var_51] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_52] = (compute_5[cse_var_52] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_53] = (compute_5[cse_var_53] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1280)], 0f32)))
+ compute_5[cse_var_54] = (compute_5[cse_var_54] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_55] = (compute_5[cse_var_55] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_56] = (compute_5[cse_var_56] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_57] = (compute_5[cse_var_57] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_58] = (compute_5[cse_var_58] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_59] = (compute_5[cse_var_59] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_60] = (compute_5[cse_var_60] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_61] = (compute_5[cse_var_61] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_63] = (compute_5[cse_var_63] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_64] = (compute_5[cse_var_64] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_65] = (compute_5[cse_var_65] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_66] = (compute_5[cse_var_66] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_67] = (compute_5[cse_var_67] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_68] = (compute_5[cse_var_68] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_69] = (compute_5[cse_var_69] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_70] = (compute_5[cse_var_70] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1536)], 0f32)))
+ compute_5[cse_var_71] = (compute_5[cse_var_71] + (placeholder_1[((placeholder_3[cse_var_2]*16) + cse_var_130)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_72] = (compute_5[cse_var_72] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 1)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_73] = (compute_5[cse_var_73] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 2)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_74] = (compute_5[cse_var_74] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 3)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_75] = (compute_5[cse_var_75] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 4)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_76] = (compute_5[cse_var_76] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 5)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_77] = (compute_5[cse_var_77] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 6)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_78] = (compute_5[cse_var_78] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 7)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_79] = (compute_5[cse_var_79] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 8)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_80] = (compute_5[cse_var_80] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 9)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_81] = (compute_5[cse_var_81] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 10)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_82] = (compute_5[cse_var_82] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 11)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_83] = (compute_5[cse_var_83] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 12)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_84] = (compute_5[cse_var_84] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 13)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_85] = (compute_5[cse_var_85] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 14)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ compute_5[cse_var_86] = (compute_5[cse_var_86] + (placeholder_1[(((placeholder_3[cse_var_2]*16) + cse_var_130) + 15)]*max(placeholder[((cse_var_131 + placeholder_2[(placeholder_3[cse_var_2] + elem_idx)]) + 1792)], 0f32)))
+ }
}
}
}
}
- for (i0.inner: int32, 0, 16) {
- let cse_var_260: int32 = (((i0.outer*8192) + (i0.inner*512)) + (i1.outer*32))
- compute[ramp(cse_var_260, 1, 32)] = max((compute_5[ramp((i0.inner*32), 1, 32)] + placeholder_4[ramp(cse_var_260, 1, 32)]), broadcast(0f32, 32))
+ for (i0.inner: int32, 0, 128) {
+ let cse_var_132: int32 = ((i0.inner*512) + (i0.outer.i1.outer.fused*32))
+ compute[ramp(cse_var_132, 1, 32)] = max((compute_5[ramp((i0.inner*32), 1, 32)] + placeholder_4[ramp(cse_var_132, 1, 32)]), broadcast(0f32, 32))
}
}
}
@@ -1448,7 +1065,7 @@ 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: 3.392 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 2.763 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 9c4ca60a4..f48d99cb8 100644
--- a/docs/how_to/tune_with_autotvm/sg_execution_times.html
+++ b/docs/how_to/tune_with_autotvm/sg_execution_times.html
@@ -327,7 +327,7 @@
<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>00:45.665</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
+<p><strong>00:46.504</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -336,11 +336,11 @@
</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>00:45.630</p></td>
+<td><p>00:46.468</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>00:00.020</p></td>
+<td><p>00:00.021</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 fb20f9c2a..b9e4ded48 100644
--- a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
+++ b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
@@ -1436,8 +1436,8 @@ No: 8 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
TimeoutError
[('tile_f', [-1, 2, 1, 64]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4909501
-No: 9 GFLOPS: 80.77/80.77 result: MeasureResult(costs=(0.0028661723142857144,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6695890426635742, timestamp=1663152848.0716567) [('tile_f', [-1, 1, 4, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5072689
-No: 10 GFLOPS: 0.00/80.77 result: Traceback (most recent call last):
+No: 9 GFLOPS: 193.79/193.79 result: MeasureResult(costs=(0.0011946226666666668,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.089510440826416, timestamp=1663169826.564597) [('tile_f', [-1, 1, 4, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5072689
+No: 10 GFLOPS: 0.00/193.79 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
@@ -1560,8 +1560,8 @@ Traceback (most recent call last):
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, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5092711
-No: 11 GFLOPS: 260.49/260.49 result: MeasureResult(costs=(0.0008887261396648044,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7274281978607178, timestamp=1663152849.0026462) [('tile_f', [-1, 8, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4264713
-No: 12 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+No: 11 GFLOPS: 260.43/260.43 result: MeasureResult(costs=(0.0008889091049723756,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7412033081054688, timestamp=1663169827.445183) [('tile_f', [-1, 8, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4264713
+No: 12 GFLOPS: 0.00/260.43 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
@@ -1684,7 +1684,7 @@ Traceback (most recent call last):
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, 1, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,183542
-No: 13 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+No: 13 GFLOPS: 0.00/260.43 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
@@ -1807,7 +1807,7 @@ Traceback (most recent call last):
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, 8, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2482196
-No: 14 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+No: 14 GFLOPS: 0.00/260.43 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
@@ -1930,9 +1930,9 @@ Traceback (most recent call last):
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, 64, 1, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10306226
-No: 15 GFLOPS: 5.25/260.49 result: MeasureResult(costs=(0.0440660225,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.857121229171753, timestamp=1663152853.5778544) [('tile_f', [-1, 2, 2, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5330964
-No: 16 GFLOPS: 3.34/260.49 result: MeasureResult(costs=(0.06935714725,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.554947137832642, timestamp=1663152854.8194664) [('tile_f', [-1, 8, 4, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2140058
-No: 17 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+No: 15 GFLOPS: 5.29/260.43 result: MeasureResult(costs=(0.04376472025000001,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8177464008331299, timestamp=1663169832.0521896) [('tile_f', [-1, 2, 2, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5330964
+No: 16 GFLOPS: 3.34/260.43 result: MeasureResult(costs=(0.06941178825,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.619394063949585, timestamp=1663169833.2994785) [('tile_f', [-1, 8, 4, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2140058
+No: 17 GFLOPS: 0.00/260.43 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 142, in build
res = future.result()
File "/usr/lib/python3.7/concurrent/futures/_base.py", line 435, in result
@@ -1950,8 +1950,8 @@ No: 17 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
TimeoutError
[('tile_f', [-1, 2, 2, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10195251
-No: 18 GFLOPS: 28.16/260.49 result: MeasureResult(costs=(0.008221106714285715,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2657346725463867, timestamp=1663152865.8246107) [('tile_f', [-1, 4, 8, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6068603
-No: 19 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+No: 18 GFLOPS: 28.23/260.43 result: MeasureResult(costs=(0.00819916792857143,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2819502353668213, timestamp=1663169844.3120673) [('tile_f', [-1, 4, 8, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6068603
+No: 19 GFLOPS: 0.00/260.43 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
@@ -2074,7 +2074,7 @@ Traceback (most recent call last):
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, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6956993
-No: 20 GFLOPS: 0.00/260.49 result: Traceback (most recent call last):
+No: 20 GFLOPS: 0.00/260.43 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
@@ -2237,7 +2237,7 @@ and measure running time.</p>
Best config:
[('tile_f', [-1, 8, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4264713
Finish loading 20 records
-Time cost of this operator: 0.001238
+Time cost of this operator: 0.001224
</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 a34bcb44a..28693f6e2 100644
--- a/docs/how_to/work_with_microtvm/micro_autotune.html
+++ b/docs/how_to/work_with_microtvm/micro_autotune.html
@@ -584,10 +584,10 @@ the tuned operator.</p>
########## Build without Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs Measurements(us)
--------- --- -------- ------- ----- ------ ------- ----------------
-tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 310.5 98.701 (1, 2, 10, 10, 3) 2 1 [310.5]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.118 0.991 (1, 6, 10, 10) 1 1 [3.118]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.969 0.308 (1, 1, 10, 10, 3) 1 1 [0.969]
-Total_time - 314.587 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 313.3 98.707 (1, 2, 10, 10, 3) 2 1 [313.3]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.121 0.983 (1, 6, 10, 10) 1 1 [3.121]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.984 0.31 (1, 1, 10, 10, 3) 1 1 [0.984]
+Total_time - 317.405 - - - - -
</pre></div>
</div>
</div>
@@ -640,10 +640,10 @@ Total_time -
########## 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 245.2 98.819 (1, 1, 10, 10, 6) 2 1 [245.2]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.955 0.788 (1, 6, 10, 10) 1 1 [1.955]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.974 0.392 (1, 1, 10, 10, 3) 1 1 [0.974]
-Total_time - 248.129 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 88.188 96.991 (1, 6, 10, 10, 1) 2 1 [88.188]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.781 1.959 (1, 6, 10, 10) 1 1 [1.781]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.955 1.051 (1, 1, 10, 10, 3) 1 1 [0.955]
+Total_time - 90.924 - - - - -
</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_tflite.html b/docs/how_to/work_with_microtvm/micro_tflite.html
index a00834dda..19d1b20d1 100644
--- a/docs/how_to/work_with_microtvm/micro_tflite.html
+++ b/docs/how_to/work_with_microtvm/micro_tflite.html
@@ -646,7 +646,7 @@ QEMU VM based on BOARD. In the example below the x86 arch is selected and a x86
extern "C"
#endif
TVM_DLL int32_t tvmgen_default_fused_nn_dense_add(void* args, int32_t* arg_type_ids, int32_t num_args, void* out_ret_value, int32_t* out_ret_tcode, void* resource_handle) {
- void* arg_placeholder = (((TVMValue*)args)[0].v_handle);
+ void* arg_p0 = (((TVMValue*)args)[0].v_handle);
- .
- ./codegen
- ./codegen/host
diff --git a/docs/how_to/work_with_microtvm/micro_train.html b/docs/how_to/work_with_microtvm/micro_train.html
index 88b1d51b7..5900605e7 100644
--- a/docs/how_to/work_with_microtvm/micro_train.html
+++ b/docs/how_to/work_with_microtvm/micro_train.html
@@ -516,7 +516,7 @@ take about <strong>2 minutes</strong> to download the Stanford Cars, while COCO
<a href="https://docs.python.org/3/library/shutil.html#shutil.move" title="shutil.move" class="sphx-glr-backref-module-shutil sphx-glr-backref-type-py-function"><span class="n">shutil</span><span class="o">.</span><span class="n">move</span></a><span class="p">(</span><span class="sa">f</span><span class="s2">"</span><span class="si">{</span><a href="https://docs.python.org/3/library/stdtypes.html#str" title="builtins.str" class="sphx-glr-backref-module-builtins sphx-glr-backref-typ [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>'/tmp/tmpy02oqjc1/images/random'
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>'/tmp/tmpqbeyr2dj/images/random'
</pre></div>
</div>
</div>
@@ -576,8 +576,8 @@ objects to other stuff? We can display some examples from our datasets using <co
<span class="n">plt</span><span class="o">.</span><span class="n">axis</span><span class="p">(</span><span class="s2">"off"</span><span class="p">)</span>
</pre></div>
</div>
-<img src="../../_images/sphx_glr_micro_train_001.png" srcset="../../_images/sphx_glr_micro_train_001.png" alt="[1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [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/tmpy02oqjc1/images/target contains 8144 images
-/tmp/tmpy02oqjc1/images/random contains 5000 images
+<img src="../../_images/sphx_glr_micro_train_001.png" srcset="../../_images/sphx_glr_micro_train_001.png" alt="[1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [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/tmpqbeyr2dj/images/target contains 8144 images
+/tmp/tmpqbeyr2dj/images/random contains 5000 images
</pre></div>
</div>
</div>
@@ -689,13 +689,13 @@ 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: 0.2103 - accuracy: 0.9256 - val_loss: 0.1451 - val_accuracy: 0.9592 - 47s/epoch - 142ms/step
+328/328 - 47s - loss: 0.2159 - accuracy: 0.9242 - val_loss: 0.1586 - val_accuracy: 0.9479 - 47s/epoch - 144ms/step
Epoch 2/3
-328/328 - 43s - loss: 0.0929 - accuracy: 0.9642 - val_loss: 0.1379 - val_accuracy: 0.9577 - 43s/epoch - 132ms/step
+328/328 - 44s - loss: 0.0987 - accuracy: 0.9630 - val_loss: 0.1163 - val_accuracy: 0.9592 - 44s/epoch - 133ms/step
Epoch 3/3
-328/328 - 43s - loss: 0.0636 - accuracy: 0.9761 - val_loss: 0.1846 - val_accuracy: 0.9460 - 43s/epoch - 132ms/step
+328/328 - 43s - loss: 0.0668 - accuracy: 0.9752 - val_loss: 0.1528 - val_accuracy: 0.9535 - 43s/epoch - 133ms/step
-<keras.callbacks.History object at 0x7f232bc3b2d0>
+<keras.callbacks.History object at 0x7ff287c2ab50>
</pre></div>
</div>
</div>
@@ -961,7 +961,7 @@ 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 25.782 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 4 minutes 29.667 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 f5163e017..d595d8825 100644
--- a/docs/how_to/work_with_microtvm/sg_execution_times.html
+++ b/docs/how_to/work_with_microtvm/sg_execution_times.html
@@ -327,7 +327,7 @@
<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>05:19.716</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
+<p><strong>05:25.473</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -336,19 +336,19 @@
</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>04:25.782</p></td>
+<td><p>04:29.667</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="micro_autotune.html#sphx-glr-how-to-work-with-microtvm-micro-autotune-py"><span class="std std-ref">Autotuning with microTVM</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_autotune.py</span></code>)</p></td>
-<td><p>00:42.520</p></td>
+<td><p>00:44.074</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><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>00:08.139</p></td>
+<td><p>00:08.282</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="micro_tflite.html#sphx-glr-how-to-work-with-microtvm-micro-tflite-py"><span class="std std-ref">microTVM with TFLite Models</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_tflite.py</span></code>)</p></td>
-<td><p>00:03.274</p></td>
+<td><p>00:03.448</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="micro_ethosu.html#sphx-glr-how-to-work-with-microtvm-micro-ethosu-py"><span class="std std-ref">Running TVM on bare metal Arm(R) Cortex(R)-M55 CPU and Ethos(TM)-U55 NPU with CMSIS-NN</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_ethosu.py</span></code>)</p></td>
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 cc606cc08..1eebc9516 100644
--- a/docs/how_to/work_with_relay/sg_execution_times.html
+++ b/docs/how_to/work_with_relay/sg_execution_times.html
@@ -327,7 +327,7 @@
<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>00:42.593</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
+<p><strong>00:44.630</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -336,19 +336,19 @@
</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>00:31.230</p></td>
+<td><p>00:32.950</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>00:09.902</p></td>
+<td><p>00:10.172</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>00:01.454</p></td>
+<td><p>00:01.501</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>
-<td><p>00:00.006</p></td>
+<td><p>00:00.007</p></td>
<td><p>0.0 MB</p></td>
</tr>
</tbody>
diff --git a/docs/how_to/work_with_schedules/intrin_math.html b/docs/how_to/work_with_schedules/intrin_math.html
index 1103dc9dd..2a304f9fb 100644
--- a/docs/how_to/work_with_schedules/intrin_math.html
+++ b/docs/how_to/work_with_schedules/intrin_math.html
@@ -522,7 +522,7 @@ The following example customizes CUDA lowering rule for <code class="code docuti
<a href="../../reference/api/python/ir.html#tvm.ir.register_intrin_lowering" title="tvm.ir.register_intrin_lowering" class="sphx-glr-backref-module-tvm-ir sphx-glr-backref-type-py-function"><span class="n">register_intrin_lowering</span></a><span class="p">(</span><span class="s2">"tir.exp"</span><span class="p">,</span> <span class="n">target</span><span class="o">=</span><span class="s2">"cuda"</span><span class="p">,</span> <span class="n">f</span><span class="o">= [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span><function my_cuda_math_rule at 0x7f235defa9e0>
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span><function my_cuda_math_rule at 0x7ff20fb4d3b0>
</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 057d3ced5..cf971d511 100644
--- a/docs/how_to/work_with_schedules/sg_execution_times.html
+++ b/docs/how_to/work_with_schedules/sg_execution_times.html
@@ -327,7 +327,7 @@
<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>00:07.049</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
+<p><strong>00:04.860</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -336,23 +336,23 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="intrin_math.html#sphx-glr-how-to-work-with-schedules-intrin-math-py"><span class="std std-ref">Intrinsics and Math Functions</span></a> (<code class="docutils literal notranslate"><span class="pre">intrin_math.py</span></code>)</p></td>
-<td><p>00:04.899</p></td>
+<td><p>00:02.419</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>00:00.937</p></td>
+<td><p>00:01.095</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>00:00.526</p></td>
+<td><p>00:00.587</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>00:00.504</p></td>
+<td><p>00:00.574</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>00:00.099</p></td>
+<td><p>00:00.101</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="schedule_primitives.html#sphx-glr-how-to-work-with-schedules-schedule-primitives-py"><span class="std std-ref">Schedule Primitives in TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">schedule_primitives.py</span></code>)</p></td>
diff --git a/docs/how_to/work_with_schedules/tensorize.html b/docs/how_to/work_with_schedules/tensorize.html
index 0f6ee7a49..9783f014a 100644
--- a/docs/how_to/work_with_schedules/tensorize.html
+++ b/docs/how_to/work_with_schedules/tensorize.html
@@ -577,7 +577,7 @@ The importing needs to happen before the tensorized GEMV being executed.</p>
C: Buffer(C_2: Pointer(float32), float32, [524288], [])}
buffer_map = {A_1: A, B_1: B, C_1: C}
preflattened_buffer_map = {A_1: A_3: Buffer(A_2, float32, [1024, 64], []), B_1: B_3: Buffer(B_2, float32, [512, 64], []), C_1: C_3: Buffer(C_2, float32, [1024, 512], [])} {
- attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmpycd4q_qg/input0.cc'\nsource_filename = \"/tmp/tmpycd4q_qg/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = allo [...]
+ attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmpzj0yygjo/input0.cc'\nsource_filename = \"/tmp/tmpzj0yygjo/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = allo [...]
for (i, 0, 1024) {
for (j.outer: int32, 0, 32) {
@tir.call_extern("gemv_update", @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), C_2, ((i*512) + (j.outer*16)), 16, 2, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), A_2, (i*64), 64, 1, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), B_2, (j.outer*1024), 1024, 1, dtype=handle), 16, 64, 64, dtype=int32)
diff --git a/docs/reference/api/python/auto_scheduler.html b/docs/reference/api/python/auto_scheduler.html
index 7aadc6d9c..4714b3aa2 100644
--- a/docs/reference/api/python/auto_scheduler.html
+++ b/docs/reference/api/python/auto_scheduler.html
@@ -1602,7 +1602,7 @@ 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>
@@ -1886,7 +1886,7 @@ 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 79f1810ef..2ed98a2c9 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/2aa0d1fbf/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L42">rpc_server.ts:42</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L63">rpc_server.ts:63</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L49">rpc_server.ts:49</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L57">rpc_server.ts:57</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 53507134a..01024b380 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/2aa0d1fbf/web/src/memory.ts#L223">memory.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/memory.ts#L223">memory.ts:223</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -172,7 +172,7 @@
<div class="tsd-signature tsd-kind-icon">temp<wbr>Args<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><a href="../interfaces/disposable.html" class="tsd-signature-type">Disposable</a><span class="tsd-signature-symbol">></span><span class="tsd-signature-symbol"> = []</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/memory.ts#L208">memory.ts:208</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L312">memory.ts:312</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L284">memory.ts:284</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L388">memory.ts:388</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L376">memory.ts:376</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L267">memory.ts:267</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L243">memory.ts:243</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L321">memory.ts:321</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L252">memory.ts:252</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L359">memory.ts:359</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L342">memory.ts:342</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L350">memory.ts:350</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L326">memory.ts:326</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L363">memory.ts:363</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L346">memory.ts:346</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L334">memory.ts:334</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 906ccae5d..92429fc1e 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/2aa0d1fbf/web/src/runtime.ts#L262">runtime.ts:262</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L260">runtime.ts:260</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L258">runtime.ts:258</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L262">runtime.ts:262</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L279">runtime.ts:279</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L270">runtime.ts:270</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 3dff46a29..eceeecb1c 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/2aa0d1fbf/web/src/runtime.ts#L202">runtime.ts:202</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L200">runtime.ts:200</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L198">runtime.ts:198</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L223">runtime.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L230">runtime.ts:230</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 4d6f77b7c..4725629bd 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/2aa0d1fbf/web/src/environment.ts#L86">environment.ts:86</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/environment.ts#L70">environment.ts:70</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/environment.ts#L70">environment.ts:70</a></li>
</ul>
</aside>
</section>
@@ -179,7 +179,7 @@
<div class="tsd-signature tsd-kind-icon">logger<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>msg<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> => </span><span class="tsd-signature-type">void</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/environment.ts#L69">environment.ts:69</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/environment.ts#L69">environment.ts:69</a></li>
</ul>
</aside>
<div class="tsd-type-declaration">
@@ -210,7 +210,7 @@
<div class="tsd-signature tsd-kind-icon">packedCFunc<wbr>Table<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">ctypes.FTVMWasmPackedCFunc</span><span class="tsd-signature-symbol"> | </span><span class="tsd-signature-type">undefined</span><span class="tsd-signature-symbol">></span><span class="tsd-signature-symbol"> = [undefined,]</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/environment.ts#L78">environment.ts:78</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/environment.ts#L78">environment.ts:78</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -228,7 +228,7 @@
<div class="tsd-signature tsd-kind-icon">packedCFunc<wbr>Table<wbr>Free<wbr>Id<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">></span><span class="tsd-signature-symbol"> = []</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/environment.ts#L84">environment.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/environment.ts#L105">environment.ts:105</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 b086f3b94..6a036cec5 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/2aa0d1fbf/web/src/runtime.ts#L49">runtime.ts:49</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/runtime.ts#L49">runtime.ts:49</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -156,7 +156,7 @@
<div class="tsd-signature tsd-kind-icon">exports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">Function</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/runtime.ts#L46">runtime.ts:46</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L45">runtime.ts:45</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L44">runtime.ts:44</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L47">runtime.ts:47</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L76">runtime.ts:76</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L66">runtime.ts:66</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L84">runtime.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L95">runtime.ts:95</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L72">runtime.ts:72</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 e70e9f3b9..3c12b94d1 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/2aa0d1fbf/web/src/runtime.ts#L583">runtime.ts:583</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L579">runtime.ts:579</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L654">runtime.ts:654</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L597">runtime.ts:597</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L631">runtime.ts:631</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L644">runtime.ts:644</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L621">runtime.ts:621</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L609">runtime.ts:609</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 160e574bd..fccf0ca94 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/2aa0d1fbf/web/src/runtime.ts#L692">runtime.ts:692</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/runtime.ts#L692">runtime.ts:692</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -202,7 +202,7 @@
<div class="tsd-signature tsd-kind-icon">exports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">Function</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/runtime.ts#L684">runtime.ts:684</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L683">runtime.ts:683</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L932">runtime.ts:932</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L994">runtime.ts:994</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L924">runtime.ts:924</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L732">runtime.ts:732</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L952">runtime.ts:952</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L816">runtime.ts:816</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L1033">runtime.ts:1033</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L846">runtime.ts:846</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L750">runtime.ts:750</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L1013">runtime.ts:1013</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L789">runtime.ts:789</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L914">runtime.ts:914</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L1145">runtime.ts:1145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L740">runtime.ts:740</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L868">runtime.ts:868</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L857">runtime.ts:857</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L940">runtime.ts:940</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 7825b43a0..6ca958083 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/2aa0d1fbf/web/src/memory.ts#L40">memory.ts:40</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L32">memory.ts:32</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L33">memory.ts:33</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L154">memory.ts:154</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L90">memory.ts:90</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L97">memory.ts:97</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L74">memory.ts:74</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L81">memory.ts:81</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L104">memory.ts:104</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L132">memory.ts:132</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L145">memory.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L60">memory.ts:60</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L67">memory.ts:67</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L53">memory.ts:53</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L114">memory.ts:114</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L124">memory.ts:124</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/memory.ts#L175">memory.ts:175</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 e387ceb1f..3106aafa7 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/2aa0d1fbf/web/src/runtime.ts#L504">runtime.ts:504</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L502">runtime.ts:502</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L516">runtime.ts:516</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L530">runtime.ts:530</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L561">runtime.ts:561</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 610ef27b0..9a7a0074b 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/2aa0d1fbf/web/src/runtime.ts#L304">runtime.ts:304</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L297">runtime.ts:297</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L293">runtime.ts:293</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L289">runtime.ts:289</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L291">runtime.ts:291</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/runtime.ts#L291">runtime.ts:291</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -218,7 +218,7 @@
<div class="tsd-signature tsd-kind-icon">shape<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/runtime.ts#L295">runtime.ts:295</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L370">runtime.ts:370</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L414">runtime.ts:414</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L355">runtime.ts:355</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L474">runtime.ts:474</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L443">runtime.ts:443</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 fd0548c00..4f6b1f631 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/2aa0d1fbf/web/src/runtime.ts#L158">runtime.ts:158</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L157">runtime.ts:157</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L165">runtime.ts:165</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 97ffb16d4..bf0496d40 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/2aa0d1fbf/web/src/rpc_server.ts#L92">rpc_server.ts:92</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/rpc_server.ts#L92">rpc_server.ts:92</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -176,7 +176,7 @@
<div class="tsd-signature tsd-kind-icon">get<wbr>Imports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> => </span><span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">unknown</span><span class="tsd-signat [...]
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/rpc_server.ts#L82">rpc_server.ts:82</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L78">rpc_server.ts:78</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/rpc_server.ts#L78">rpc_server.ts:78</a></li>
</ul>
</aside>
</section>
@@ -211,7 +211,7 @@
<div class="tsd-signature tsd-kind-icon">logger<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>msg<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> => </span><span class="tsd-signature-type">void</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/2aa0d1fbf/web/src/rpc_server.ts#L81">rpc_server.ts:81</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L79">rpc_server.ts:79</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L80">rpc_server.ts:80</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/rpc_server.ts#L77">rpc_server.ts:77</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 de32cc1ae..f3d84dd85 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/2aa0d1fbf/web/src/runtime.ts#L145">runtime.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L145">runtime.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L143">runtime.ts:143</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 8f4bf2290..56283677d 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/2aa0d1fbf/web/src/webgpu.ts#L57">webgpu.ts:57</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/webgpu.ts#L50">webgpu.ts:50</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/webgpu.ts#L51">webgpu.ts:51</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/webgpu.ts#L84">webgpu.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/webgpu.ts#L170">webgpu.ts:170</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/webgpu.ts#L67">webgpu.ts:67</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 d343528a0..4a5e5f9f5 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/2aa0d1fbf/web/src/ctypes.ts#L220">ctypes.ts:220</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L216">ctypes.ts:216</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L214">ctypes.ts:214</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L218">ctypes.ts:218</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L226">ctypes.ts:226</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L221">ctypes.ts:221</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L219">ctypes.ts:219</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L223">ctypes.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L227">ctypes.ts:227</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L222">ctypes.ts:222</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L228">ctypes.ts:228</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L217">ctypes.ts:217</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L224">ctypes.ts:224</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L225">ctypes.ts:225</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/ctypes.ts#L215">ctypes.ts:215</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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 1c8a61b02..b06d495fa 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/2aa0d1fbf/web/src/runtime.ts#L676">runtime.ts:676</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/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/2aa0d1fbf/web/src/runtime.ts#L675">runtime.ts:675</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/a40849342/web/src/runtime.ts#L675">runtime.ts:675</a></li>
... 1522 lines suppressed ...