Merge branch 'develop' into migx-device-interface

docs/source/Linux_Install_Guide.rst

-=====================
-Getting Started Guide
-=====================
-
-------------
-Introduction
-------------
-
-This document contains instructions for installing, using, and contributing to Composable Kernel (CK).
-
-Documentation Roadmap
-^^^^^^^^^^^^^^^^^^^^^
-The following is a list of CK documents in the suggested reading order:
-
-[TODO]
\ No newline at end of file

docs/source/Makefile

-# Minimal makefile for Sphinx documentation
-#
-
-# You can set these variables from the command line.
-SPHINXOPTS    =
-SPHINXBUILD   = sphinx-build
-SPHINXPROJ    = CK
-SOURCEDIR     = .
-BUILDDIR      = _build
-
-# Put it first so that "make" without argument is like "make help".
-help:
-	@$(SPHINXBUILD) -M help "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)
-
-.PHONY: help Makefile
-
-# Catch-all target: route all unknown targets to Sphinx using the new
-# "make mode" option.  $(O) is meant as a shortcut for $(SPHINXOPTS).
-%: Makefile
-	@$(SPHINXBUILD) -M $@ "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)

docs/source/conf.py

-"""Copyright (C) 2018-2023 Advanced Micro Devices, Inc. All rights reserved.
-
-   Permission is hereby granted, free of charge, to any person obtaining a copy
-   of this software and associated documentation files (the "Software"), to deal
-   in the Software without restriction, including without limitation the rights
-   to use, copy, modify, merge, publish, distribute, sublicense, and/or sell cop-
-   ies of the Software, and to permit persons to whom the Software is furnished
-   to do so, subject to the following conditions:
-
-   The above copyright notice and this permission notice shall be included in all
-   copies or substantial portions of the Software.
-
-   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IM-
-   PLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
-   FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
-   COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
-   IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNE-
-   CTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-"""
-
-# -*- coding: utf-8 -*-
-#
-# Composable Kernel (CK) docuumentation build configuration file, based on
-# rocBLAS documentation build configuration file, created by
-# sphinx-quickstart on Mon Jan  8 16:34:42 2018.
-#
-# This file is execfile()d with the current directory set to its
-# containing dir.
-#
-# Note that not all possible configuration values are present in this
-# autogenerated file.
-#
-# All configuration values have a default; values that are commented out
-# serve to show the default.
-
-# If extensions (or modules to document with autodoc) are in another directory,
-# add these directories to sys.path here. If the directory is relative to the
-# documentation root, use os.path.abspath to make it absolute, like shown here.
-#
-# import os
-# import sys
-# sys.path.insert(0, os.path.abspath('.'))
-
-import os
-import sys
-import subprocess
-
-read_the_docs_build = os.environ.get('READTHEDOCS', None) == 'True'
-
-if read_the_docs_build:
-    subprocess.call('../run_doxygen.sh')
-
-# -- General configuration ------------------------------------------------
-
-# If your documentation needs a minimal Sphinx version, state it here.
-#
-# needs_sphinx = '1.0'
-
-# Add any Sphinx extension module names here, as strings. They can be
-# extensions coming with Sphinx (named 'sphinx.ext.*') or your custom
-# ones.
-extensions = ['sphinx.ext.mathjax', 'breathe', 'sphinxcontrib.bibtex']
-
-breathe_projects = { "CK": "../docBin/xml" }
-breathe_default_project = "CK"
-
-bibtex_bibfiles = ['refs.bib']
-
-# Add any paths that contain templates here, relative to this directory.
-templates_path = ['_templates']
-
-# The suffix(es) of source filenames.
-# You can specify multiple suffix as a list of string:
-#
-# source_suffix = ['.rst', '.md']
-source_suffix = '.rst'
-
-# The master toctree document.
-master_doc = 'index'
-
-# General information about the project.
-project = u'Composable Kernel (CK)'
-copyright = u'2018-2023, Advanced Micro Devices'
-author = u'Advanced Micro Devices'
-
-# The version info for the project you're documenting, acts as replacement for
-# |version| and |release|, also used in various other places throughout the
-# built documents.
-#
-# The short X.Y version.
-#version = u'0.8'
-# The full version, including alpha/beta/rc tags.
-#release = u'0.8'
-
-# The language for content autogenerated by Sphinx. Refer to documentation
-# for a list of supported languages.
-#
-# This is also used if you do content translation via gettext catalogs.
-# Usually you set "language" from the command line for these cases.
-language = 'en'
-
-# List of patterns, relative to source directory, that match files and
-# directories to ignore when looking for source files.
-# This patterns also effect to html_static_path and html_extra_path
-exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store']
-
-# The name of the Pygments (syntax highlighting) style to use.
-pygments_style = 'sphinx'
-
-# If true, `todo` and `todoList` produce output, else they produce nothing.
-todo_include_todos = False
-
-
-# -- Options for HTML output ----------------------------------------------
-
-# The theme to use for HTML and HTML Help pages.  See the documentation for
-# a list of builtin themes.
-#
-# html_theme = 'alabaster'
-
-#if read_the_docs_build:
-#    html_theme = 'default'
-#else:
-import sphinx_rtd_theme
-html_theme = "sphinx_rtd_theme"
-html_theme_path = [sphinx_rtd_theme.get_html_theme_path()]
-html_logo = "rocm_logo.png"
-
-# Theme options are theme-specific and customize the look and feel of a theme
-# further.  For a list of options available for each theme, see the
-# documentation.
-html_theme_options = {
-    'logo_only': True,
-    'display_version': True
-}
-
-# Add any paths that contain custom static files (such as style sheets) here,
-# relative to this directory. They are copied after the builtin static files,
-# so a file named "default.css" will overwrite the builtin "default.css".
-#html_static_path = ['_static']
-
-# Custom sidebar templates, must be a dictionary that maps document names
-# to template names.
-#
-# This is required for the alabaster theme
-# refs: http://alabaster.readthedocs.io/en/latest/installation.html#sidebars
-# html_sidebars = {
-#     '**': [
-#         'relations.html',  # needs 'show_related': True theme option to display
-#         'searchbox.html',
-#     ]
-# }
-
-mathjax3_config = {
-'tex': {
-    'macros': {
-        'diag': '\\operatorname{diag}',
-        }
-    }
-}
-
-# -- Options for HTMLHelp output ------------------------------------------
-
-# Output file base name for HTML help builder.
-htmlhelp_basename = 'CKdoc'
-
-
-# -- Options for LaTeX output ---------------------------------------------
-
-latex_elements = {
-    # The paper size ('letterpaper' or 'a4paper').
-    #
-    'papersize': 'letterpaper',
-
-    # The font size ('10pt', '11pt' or '12pt').
-    #
-    'pointsize': '10pt',
-
-    # Additional stuff for the LaTeX preamble.
-    #
-    'preamble': r'''
-\setcounter{tocdepth}{5}
-\newcommand{\diag}{\operatorname{diag}}
-''',
-
-    # Latex figure (float) alignment
-    #
-    # 'figure_align': 'htbp',
-}
-
-# Grouping the document tree into LaTeX files. List of tuples
-# (source start file, target name, title,
-#  author, documentclass [howto, manual, or own class]).
-latex_documents = [
-    (master_doc, 'CK.tex', u'Composabl Kernel (CK) Documentation',
-     u'Advanced Micro Devices', 'manual'),
-]
-
-
-# -- Options for manual page output ---------------------------------------
-
-# One entry per manual page. List of tuples
-# (source start file, name, description, authors, manual section).
-man_pages = [
-    (master_doc, 'ck', u'Composable Kernel (CK) Documentation',
-     [author], 1)
-]
-
-
-# -- Options for Texinfo output -------------------------------------------
-
-# Grouping the document tree into Texinfo files. List of tuples
-# (source start file, target name, title, author,
-#  dir menu entry, description, category)
-texinfo_documents = [
-    (master_doc, 'CK', u'Composable Kernel (CK) Documentation',
-     author, 'CK', 'Composable Kernel for AMD ROCm',
-     'Miscellaneous'),
-]

docs/source/index.rst

-============================
-Composable Kernel User Guide
-============================
-
-.. toctree::
-   :maxdepth: 5
-   :caption: Contents:
-   :numbered:
-
-   Linux_Install_Guide
-   tutorial_hello_world
-   dockerhub
-   Supported_Primitives_Guide
-   API_Reference_Guide
-   Contributors_Guide
-   Disclaimer
\ No newline at end of file

example/30_grouped_conv_fwd_multiple_d/grouped_conv_fwd_xdl_fp16.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
 

example/30_grouped_conv_fwd_multiple_d/run_grouped_conv_fwd_bias_relu_add_wmma_example.inc

@@ -74,8 +74,8 @@ using DeviceConvFwdInstance =
         8,           // BBlockTransferSrcScalarPerVector
         8,           // BBlockTransferDstScalarPerVector_BK1
         true,        // BBlockLdsExtraN
-        1,
-        1,
+        4,
+        2,
         S<1, 32, 1, 8>,
         8>;
 

example/40_conv2d_fwd_quantization/CMakeLists.txt

@@ -14,3 +14,8 @@ add_example_executable(example_conv2d_fwd_xdl_bias_relu_perlayer_quantization_in
 add_example_executable(example_conv2d_fwd_dl_bias_relu_perchannel_quantization_int8 conv2d_fwd_dl_bias_relu_perchannel_quantization_int8.cpp)
 add_example_executable(example_conv2d_fwd_xdl_bias_relu_perchannel_quantization_int8 conv2d_fwd_xdl_bias_relu_perchannel_quantization_int8.cpp)
 
+# Conv + bias + tanh perlayer quantization
+add_example_executable(example_conv2d_fwd_dl_bias_tanh_perlayer_quantization_int8 conv2d_fwd_dl_bias_tanh_perlayer_quantization_int8.cpp)
+
+# Conv + bias + tanh perchannel quantization
+add_example_executable(example_conv2d_fwd_dl_bias_tanh_perchannel_quantization_int8 conv2d_fwd_dl_bias_tanh_perchannel_quantization_int8.cpp)

example/40_conv2d_fwd_quantization/conv2d_fwd_dl_bias_relu_perchannel_quantization_int8.cpp

@@ -76,6 +76,10 @@ using DeviceGroupedConvNDFwdInstance =
         5,                   // CThreadTransferSrcDstVectorDim
         4>;                  // CThreadTransferDstScalarPerVector
 
-#include "run_conv2d_fwd_bias_relu_perchannel_quantization_example.inc"
+#include "run_conv2d_fwd_bias_perchannel_quantization_example.inc"
 
-int main() { run_conv2d_fwd_bias_relu_perchannel_quantization_example(); };
+int main()
+{
+    const auto out_element_op = OutElementOp{ActivationOp{}};
+    run_conv2d_fwd_bias_perchannel_quantization_example(out_element_op);
+};

example/40_conv2d_fwd_quantization/conv2d_fwd_dl_bias_relu_perlayer_quantization_int8.cpp

@@ -74,6 +74,11 @@ using DeviceGroupedConvNDFwdInstance =
         5,                   // CThreadTransferSrcDstVectorDim
         4>;                  // CThreadTransferDstScalarPerVector
 
-#include "run_conv2d_fwd_bias_relu_perlayer_quantization_example.inc"
+#include "run_conv2d_fwd_bias_perlayer_quantization_example.inc"
 
-int main() { run_conv2d_fwd_bias_relu_perlayer_quantization_example(); }
+int main()
+{
+    float requant_scale       = 0.5f;
+    const auto out_element_op = OutElementOp{requant_scale, ActivationOp{}};
+    run_conv2d_fwd_bias_perlayer_quantization_example(out_element_op);
+}

example/40_conv2d_fwd_quantization/conv2d_fwd_dl_bias_tanh_perchannel_quantization_int8.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "common.hpp"
+#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_dl_multiple_d_nhwc_kyxc_nhwk.hpp"
+
+using InDataType           = int8_t;
+using WeiDataType          = int8_t;
+using BiasDataType         = int32_t;
+using RequantScaleDataType = float;
+using AccDataType          = int32_t;
+using OutDataType          = int8_t;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough  = ck::tensor_operation::element_wise::PassThrough;
+using InElementOp  = PassThrough;
+using WeiElementOp = PassThrough;
+using ActivationOp = ck::tensor_operation::element_wise::TanH;
+using OutElementOp =
+    ck::tensor_operation::element_wise::Add_Mul2_Activation_Mul_Clamp<ActivationOp>;
+
+static constexpr auto ConvSpec =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
+
+static constexpr auto GemmSpec = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
+
+template <ck::index_t NDimSpatial,
+          typename InLayout,
+          typename WeiLayout,
+          typename BiasLayout,
+          typename RequantScaleLayout,
+          typename OutLayout>
+using DeviceGroupedConvNDFwdInstance =
+    ck::tensor_operation::device::DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK<
+        NDimSpatial,
+        InDataType,
+        WeiDataType,
+        ck::Tuple<BiasDataType, RequantScaleDataType>,
+        OutDataType,
+        AccDataType,
+        InLayout,
+        WeiLayout,
+        ck::Tuple<BiasLayout, RequantScaleLayout>,
+        OutLayout,
+        InElementOp,
+        WeiElementOp,
+        OutElementOp,
+        ConvSpec,            // ConvForwardSpecialization
+        GemmSpec,            // GemmSpecialization
+        256,                 // BlockSize
+        128,                 // MPerBlock
+        128,                 // NPerBlock
+        16,                  // K0PerBlock
+        4,                   // K1
+        4,                   // M1PerThread
+        4,                   // N1PerThread
+        1,                   // KPerThread
+        S<8, 2>,             // M1N1ThreadClusterM1Xs
+        S<8, 2>,             // M1N1ThreadClusterN1Xs
+        S<8, 1, 1, 4>,       // ABlockTransferThreadSliceLengths_K0_M0_M1_K1
+        S<2, 1, 128, 1>,     // ABlockTransferThreadClusterLengths_K0_M0_M1_K1
+        S<1, 2, 0, 3>,       // ABlockTransferThreadClusterArrangeOrder
+        S<1, 2, 0, 3>,       // ABlockTransferSrcAccessOrder
+        S<4, 1, 1, 4>,       // ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1
+        S<1, 2, 0, 3>,       // ABlockTransferSrcVectorTensorContiguousDimOrder
+        S<1, 1, 1, 4>,       // ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1
+        S<8, 1, 1, 4>,       // BBlockTransferThreadSliceLengths_K0_N0_N1_K1
+        S<2, 1, 128, 1>,     // BBlockTransferThreadClusterLengths_K0_N0_N1_K1
+        S<1, 2, 0, 3>,       // BBlockTransferThreadClusterArrangeOrder
+        S<1, 2, 0, 3>,       // BBlockTransferSrcAccessOrder
+        S<4, 1, 1, 4>,       // BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1
+        S<1, 2, 0, 3>,       // BBlockTransferSrcVectorTensorContiguousDimOrder
+        S<1, 1, 1, 4>,       // BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1
+        S<0, 1, 2, 3, 4, 5>, // CThreadTransferSrcDstAccessOrder
+        5,                   // CThreadTransferSrcDstVectorDim
+        4>;                  // CThreadTransferDstScalarPerVector
+
+#include "run_conv2d_fwd_bias_perchannel_quantization_example.inc"
+
+int main()
+{
+    float scale_z_inv         = 0.5f;
+    const auto out_element_op = OutElementOp{scale_z_inv, ActivationOp{}};
+    run_conv2d_fwd_bias_perchannel_quantization_example(out_element_op);
+};

example/40_conv2d_fwd_quantization/conv2d_fwd_dl_bias_tanh_perlayer_quantization_int8.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "common.hpp"
+#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_dl_multiple_d_nhwc_kyxc_nhwk.hpp"
+
+using InDataType   = int8_t;
+using WeiDataType  = int8_t;
+using BiasDataType = int32_t;
+using AccDataType  = int32_t;
+using OutDataType  = int8_t;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough  = ck::tensor_operation::element_wise::PassThrough;
+using InElementOp  = PassThrough;
+using WeiElementOp = PassThrough;
+using ActivationOp = ck::tensor_operation::element_wise::TanH;
+using OutElementOp = ck::tensor_operation::element_wise::Add_Mul_Activation_Mul_Clamp<ActivationOp>;
+
+static constexpr auto ConvSpec =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
+
+static constexpr auto GemmSpec = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
+
+template <ck::index_t NDimSpatial,
+          typename InLayout,
+          typename WeiLayout,
+          typename BiasLayout,
+          typename OutLayout>
+using DeviceGroupedConvNDFwdInstance =
+    ck::tensor_operation::device::DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK<
+        NDimSpatial,
+        InDataType,
+        WeiDataType,
+        ck::Tuple<BiasDataType>,
+        OutDataType,
+        AccDataType,
+        InLayout,
+        WeiLayout,
+        ck::Tuple<BiasLayout>,
+        OutLayout,
+        InElementOp,
+        WeiElementOp,
+        OutElementOp,
+        ConvSpec,            // ConvForwardSpecialization
+        GemmSpec,            // GemmSpecialization
+        256,                 // BlockSize
+        128,                 // MPerBlock
+        128,                 // NPerBlock
+        16,                  // K0PerBlock
+        4,                   // K1
+        4,                   // M1PerThread
+        4,                   // N1PerThread
+        1,                   // KPerThread
+        S<8, 2>,             // M1N1ThreadClusterM1Xs
+        S<8, 2>,             // M1N1ThreadClusterN1Xs
+        S<8, 1, 1, 4>,       // ABlockTransferThreadSliceLengths_K0_M0_M1_K1
+        S<2, 1, 128, 1>,     // ABlockTransferThreadClusterLengths_K0_M0_M1_K1
+        S<1, 2, 0, 3>,       // ABlockTransferThreadClusterArrangeOrder
+        S<1, 2, 0, 3>,       // ABlockTransferSrcAccessOrder
+        S<4, 1, 1, 4>,       // ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1
+        S<1, 2, 0, 3>,       // ABlockTransferSrcVectorTensorContiguousDimOrder
+        S<1, 1, 1, 4>,       // ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1
+        S<8, 1, 1, 4>,       // BBlockTransferThreadSliceLengths_K0_N0_N1_K1
+        S<2, 1, 128, 1>,     // BBlockTransferThreadClusterLengths_K0_N0_N1_K1
+        S<1, 2, 0, 3>,       // BBlockTransferThreadClusterArrangeOrder
+        S<1, 2, 0, 3>,       // BBlockTransferSrcAccessOrder
+        S<4, 1, 1, 4>,       // BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1
+        S<1, 2, 0, 3>,       // BBlockTransferSrcVectorTensorContiguousDimOrder
+        S<1, 1, 1, 4>,       // BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1
+        S<0, 1, 2, 3, 4, 5>, // CThreadTransferSrcDstAccessOrder
+        5,                   // CThreadTransferSrcDstVectorDim
+        4>;                  // CThreadTransferDstScalarPerVector
+
+#include "run_conv2d_fwd_bias_perlayer_quantization_example.inc"
+
+int main()
+{
+    float scale_acc           = 0.5f;
+    float scale_z_inv         = 0.5f;
+    const auto out_element_op = OutElementOp{scale_z_inv, scale_acc, ActivationOp{}};
+    run_conv2d_fwd_bias_perlayer_quantization_example(out_element_op);
+}

example/40_conv2d_fwd_quantization/conv2d_fwd_dl_perchannel_quantization_int8.cpp

@@ -76,4 +76,8 @@ using DeviceGroupedConvNDFwdInstance =
 
 #include "run_conv2d_fwd_perchannel_quantization_example.inc"
 
-int main() { run_conv2d_fwd_perchannel_quantization_example(); }
+int main()
+{
+    const auto out_element_op = OutElementOp{ActivationOp{}};
+    run_conv2d_fwd_perchannel_quantization_example(out_element_op);
+}

example/40_conv2d_fwd_quantization/conv2d_fwd_dl_perlayer_quantization_int8.cpp

@@ -71,4 +71,9 @@ using DeviceGroupedConvNDFwdInstance =
 
 #include "run_conv2d_fwd_perlayer_quantization_example.inc"
 
-int main() { run_conv2d_fwd_perlayer_quantization_example(); }
+int main()
+{
+    float requant_scale       = 0.5f;
+    const auto out_element_op = OutElementOp{requant_scale, ActivationOp{}};
+    run_conv2d_fwd_perlayer_quantization_example(out_element_op);
+}

example/40_conv2d_fwd_quantization/conv2d_fwd_xdl_bias_relu_perchannel_quantization_int8.cpp

@@ -80,6 +80,10 @@ using DeviceGroupedConvNDFwdInstance =
         S<1, 64, 1, 4>,
         8>;
 
-#include "run_conv2d_fwd_bias_relu_perchannel_quantization_example.inc"
+#include "run_conv2d_fwd_bias_perchannel_quantization_example.inc"
 
-int main() { run_conv2d_fwd_bias_relu_perchannel_quantization_example(); };
+int main()
+{
+    const auto out_element_op = OutElementOp{ActivationOp{}};
+    run_conv2d_fwd_bias_perchannel_quantization_example(out_element_op);
+};

example/40_conv2d_fwd_quantization/conv2d_fwd_xdl_bias_relu_perlayer_quantization_int8.cpp

@@ -78,6 +78,11 @@ using DeviceGroupedConvNDFwdInstance =
         S<1, 64, 1, 4>,
         8>;
 
-#include "run_conv2d_fwd_bias_relu_perlayer_quantization_example.inc"
+#include "run_conv2d_fwd_bias_perlayer_quantization_example.inc"
 
-int main() { run_conv2d_fwd_bias_relu_perlayer_quantization_example(); }
+int main()
+{
+    float requant_scale       = 0.5f;
+    const auto out_element_op = OutElementOp{requant_scale, ActivationOp{}};
+    run_conv2d_fwd_bias_perlayer_quantization_example(out_element_op);
+}

example/40_conv2d_fwd_quantization/conv2d_fwd_xdl_perchannel_quantization_int8.cpp

@@ -80,4 +80,8 @@ using DeviceGroupedConvNDFwdInstance =
 
 #include "run_conv2d_fwd_perchannel_quantization_example.inc"
 
-int main() { run_conv2d_fwd_perchannel_quantization_example(); }
+int main()
+{
+    const auto out_element_op = OutElementOp{ActivationOp{}};
+    run_conv2d_fwd_perchannel_quantization_example(out_element_op);
+}

example/40_conv2d_fwd_quantization/conv2d_fwd_xdl_perlayer_quantization_int8.cpp

@@ -75,4 +75,9 @@ using DeviceGroupedConvNDFwdInstance =
 
 #include "run_conv2d_fwd_perlayer_quantization_example.inc"
 
-int main() { run_conv2d_fwd_perlayer_quantization_example(); }
+int main()
+{
+    float requant_scale       = 0.5f;
+    const auto out_element_op = OutElementOp{requant_scale, ActivationOp{}};
+    run_conv2d_fwd_perlayer_quantization_example(out_element_op);
+}

example/40_conv2d_fwd_quantization/run_conv2d_fwd_bias_relu_perchannel_quantization_example.inc → example/40_conv2d_fwd_quantization/run_conv2d_fwd_bias_perchannel_quantization_example.inc

@@ -167,7 +167,7 @@ bool run_grouped_conv_fwd(bool do_verification,
     return (pass ? 0 : 1);
 }
 
-int run_conv2d_fwd_bias_relu_perchannel_quantization_example()
+int run_conv2d_fwd_bias_perchannel_quantization_example(const OutElementOp& out_element_op)
 {
     bool do_verification           = true;
     bool time_kernel               = true;
@@ -189,7 +189,6 @@ int run_conv2d_fwd_bias_relu_perchannel_quantization_example()
 
     const auto in_element_op  = InElementOp{};
     const auto wei_element_op = WeiElementOp{};
-    const auto out_element_op = OutElementOp{ActivationOp{}};
 
     using InLayout           = ck::tensor_layout::convolution::GNHWC;
     using WeiLayout          = ck::tensor_layout::convolution::GKYXC;