Skip to content

GitLab

Switch branch/tag
History Find file
Clone
disable fp64 conv example as test
Chao Liu authored
e1db59ae
Name Last commit Last update
cmake Add host API (#220)
example disable fp64 conv example as test
external/include/half Reorganize files, Part 1 (#119)
include/ck fix
library Add FP64 XDL GEMM built-in function (#199)
profiler Add FP64 XDL GEMM built-in function (#199)
script Overhaul to Reducton and its dependants (#237)
test Add FP64 XDL GEMM built-in function (#199)
.clang-format start adding convolution
.clang-tidy add tidy
.gitignore Allow distinct K0/K1 values for A/B block descriptor (#98)
CMakeLists.txt remove options.hpp.in (#240)
Config.cmake.in Add host API (#220)
Dockerfile Add performance tests as a stage of CI. (#247)
Jenkinsfile Add performance tests as a stage of CI. (#247)
README.md Add host API (#220)
dev-requirements.txt Initial Setup for CI (#86)
rbuild.ini Update test CMakeLists to add new tests automatically and add Jenkins stage for tests (#88)
requirements.txt Update test CMakeLists to add new tests automatically and add Jenkins stage for tests (#88)
README.md

Docker script

docker run                                     \
-it                                            \
--privileged                                   \
--group-add sudo                               \
-w /root/workspace                             \
-v ${PATH_TO_LOCAL_WORKSPACE}:/root/workspace  \
rocm/tensorflow:rocm4.3.1-tf2.6-dev            \
/bin/bash

Build

mkdir build && cd build
# Need to specify target ID, example below is gfx908 and gfx90a
cmake                                                                 \
-D BUILD_DEV=OFF                                                      \
-D CMAKE_BUILD_TYPE=Release                                           \
-D CMAKE_CXX_FLAGS=" --offload-arch=gfx908 --offload-arch=gfx90a -O3" \
-D CMAKE_CXX_COMPILER=/opt/rocm/bin/hipcc                             \
-D CMAKE_PREFIX_PATH=/opt/rocm                                        \
..

Build and Run Examples

 make -j examples

Instructions for running each individual examples are under example/

Tests

 make -j tests
 make test

Build ckProfiler

 make -j ckProfiler

Instructions for running ckProfiler are under profiler/

Caveat

Kernel Timing and Verification

CK's own kernel timer will warn up kernel once, and then run it multiple times to get average kernel time. For some kernels that use atomic add, this will cause output buffer to be accumulated multiple times, causing verfication failure. To work around it, do not use CK's own timer and do verification at the same time. CK's own timer and verification in each example and ckProfiler can be enabled or disabled from command line.