GitLab

* Add flag to enable cpu backend

* Make buffers shared

* Enable optimizations

* Add onednn

* Formatting

* Formatting

* Add dnnl header

* Formatting

* Rewrite rnn first

* Formatting

* Call reference implementation

* Formatting

* Make literal data shared

* Formatting

* Add convolution

* Formatting

* Compensate for dilation

* Formatting

* Use name/make_op instead

* Formatting

* Rename gemm header

* Formatting

* Add dnnl convolution/gemm operators

* Formatting

* Add eliminate_contiguous

* Add faster pointwise operators

* Formatting

* Formatting

* Formatting

* Add dnnl op class

* Formatting

* Add add op

* Formatting

* Add concat operator

* Formatting

* Add more ops

* Create descriptor during finalization

* Formatting

* Dont rewrite pooling

* Enable memory coloring

* Formatting

* Add output aliases

* Formatting

* Fix errors

* Formatting

* Convert literals

* Add missing file

* Remove batch_norm

* Formatting

* Use strides

* Formatting

* Add some debug checks

* Formatting

* Fix big in adjusting shape for gemm

* Formatting

* Fix fallback dot operator

* Zero initialize buffers

* Add suport for group convolutions

* Formatting

* Make adjust allocation target independent

* Formatting

* Enable adjust_allocation for gpu/cpu

* Formatting

* Add copy to allocation model

* Formatting

* Add copy operator

* Formatting

* Better handling of output parameters in adjust_allocation

* Formatting

* Build with dnnl

* Make dnnl required

* Fix compile error

* Tidy fixes

* Formatting

* Tidy fixes

* Formatting

* Fix more tidy issues

* Formatting

* Add mul op

* Add mul op

* Set c compiler to clang as well

* Compensate for normalized compute shape

* Formatting

* Fix cppcheck errors

* Formatting

* Add onednn library to hcc

* Guard clang pragmas

* Disable cpu mode for gcc for now

* Leave it enabled it for gcc 7

* Fix cppcheck suppresion

* Fix compile error on gcc 5

* Remove unused code
Co-authored-by: Shucai Xiao <shucai.xiao@amd.com>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* Add compiler flags

* Add missing include

* Add filesystem header

* Formatting

* Add tmp_dir to run

* Formatting

* Kernel compilation and launching

* Formatting

* Seperate pack_args

* Formatting

* Add alignment tests

* Formatting

* Add compile test

* Formatting

* Complete compile test

* Formatting

* Use is_regular_file free function

* Fix is_regular_file call

* Fix tidy issues

* Fix tidy

* Fix tidy issue

* Print size in read_buffer to debug issue on jenkins

* Add hip flags before src file

* Fix reading output files

* Fix unsued variable warning

* Formatting

* Formatting

* Disable tidy check
Co-authored-by: Shucai Xiao <shucai.xiao@amd.com>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* Added greater and less operators

* Fixed ops_test.cpp

* Set commutative to false for less, greater

* Refactored parse_equal/less/greater into parse_compare_op

* Removed unnecessary function attributes() from greater.hpp/less.hpp

* Added op_name arguments

* Removed local settings

* Formatting

* Missing comma

* Formatting

* Formatting

* Formatting

* Formatting

* Formatting

* Missing space
Co-authored-by: Paul Fultz II <pfultz2@yahoo.com>

* Add intial multi stream analysis

* Formatting

* Add more tests

* Formatting

* Remove comment

* Analyze streams on the gpu

* Formatting

* Fix nstream

* Formatting

* Add test for return

* Formatting

* Make sure return has a stream assignment

* Formatting

* Fix asserts and checks

* Improve error message for out-of-order sequence

* Formatting

* Make global variables const

* Tidy fixes

* Disable some lints

* Formatting

* Fix tidy const

* Formatting

* Add missing const keywords

* Formatting

* More fixes

* Fix remaining tidy issues

* Formatting

* Fix rocblas function call

* Formatting

* Fix nodiscard warnings

* Formatting

* Use named parameters

* Remove overload

* Add overload

* Remove noncps

* Use named param for node

* Add auto register header

* Use named parameters

* Refactor jenkinsfile

* Fix shadow

* Add missing body variable

* Add more const methods

* Add hip-clang docker builds

* Remove comments

* Add clang-format

* Add more const

* Formatting

* Rename stage

* Disable check

* Add another const

* Add python 2 dev packages

* Add sphinx to dockerfile

* add bool type

* code backup

* code backup

* clang format

* fix build warnings

* clang format

* add the equal operator

* add the equal operator

* clang format

* remove unnecessary code

* refine unit tests

* clang format

* fix review comments and a bug

* clang format

* additional changes

* clang format

* fix cppcheck error

* add bool type in c api

* fix cppcheck error

* fix review comments

* fix cppcheck error

* fix a build error related to gcc

* fix cppcheck error

* fix cppcheck error

* added the equal operator to register list

* add parsing boolean type

* clang format

* fix bool type issue for python output

* clang format

* add support for automatic multibroadcast of the equal operator

* additional unit tests for more code coverage

* clang format

* missing an onnx file
Co-authored-by: Paul Fultz II <pfultz2@yahoo.com>

* Register ops for main migraphx

* Formatting

* Register cpu ops

* Formatting

* Show list of operators in the driver

* Formatting

* Simplify regiter

* Try to register gpu ops

* Fix compiler errors

* Register rest of the gpu operators

* Add some tests

* Formatting

* Fix gcc compiler warnings

* Formatting

* Fix tidy warnings

* Fix compile error

* Use correct op name

* Register layer norm

* Use const ref

* Make run const

* fix pad calc

* bert tf passes correctness

* formatting

* add test

* formatting

* remove comment

* add inline

* formatting

* fix order for literal

* formatting

* test no mul_add

* formatting

* debug layernorm

* debug layernorm

* manual merge

* more progress

* formatting

* remove miopen batchnorm

* remove headers

* Fix compile error with no dpp reductions

* fix indices

* formatting

* change matcher

* formatting

* remove binds

* formatting

* disable tf matcher

* formatting

* use fast div

* formatting

* fix matcher

* formatting

* remove comment

* move find_matches

* add assert

* formatting

* fix deepcode issue
Co-authored-by: Paul <pfultz2@yahoo.com>
Co-authored-by: Shucai Xiao <shucai.xiao@amd.com>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* initial progress

* formatting

* add pooling changes

* formatting

* change eliminate_pad

* formatting

* rename var

* fomratting

* update op shape test and compute

* formatting

* revert conv constructor

* formatting

* change initializer

* formatting

* fix tidy

* change quant conv and shape check

* add tests and fixes

* formatting

* fix type

* fix conv test

* formatting

* add pooling and bn tests

* formatting

* add inconsistent attr tests

* fix padding issue

* formatting

* progress on 1d to 2d

* formatting

* change compute and compile functions

* formatting

* fix duplicate

* fix conflict

* fix issue with 1d conv

* formatting

* add check for 3d limit

* rename function

* formatting

* update to MIOPen 2.3

* add support for nd pooling

* formatting

* test miopen 2.4

* change function name

* rename functions

* formatting

* add op_shape test

* add gpu ops tests

* formatting

* initial progress

* formatting

* add pkg-config

* add to support asymmetric padding of averagepool

* clang format

* fix bug for average pooling

* clang format

* fix a bug

* add unit tests for the asymmetric padding of averagepool

* clang format

* change functions

* formatting

* additional code refinement

* clang format

* check existing tests

* formatting

* change to copy_backward

* formatting

* change for loop to transform

* formatting

* add tests

* formatting

* remove comment

* add more tests

* remove an optimization for pooling

* clang format

* add and fix unit tests

* clang format

* update gpu miopen calls

* formatting

* initial progress

* add cpu impl and tests

* formatting

* add NOLINT

* add 3d test

* formatting

* add more op_shape tests

* test diff miopen version

* add submodule onnx

* add pooling shape tests

* fix error msg

* add onnx_test_backend

* reorganize python code

* temp disable test

* fix cppcheck error

* fix cppcheck error

* code backup

* add support device choice

* refine onnx backend test

* revert to miopen 2.4

* fix review comments

* fix review comments

* clang format

* fixed review comments

* clang format

* fix cppcheck error

* copy onnx_backend_test to dest when building

* add testdata folder

* fix bounds

* formatting

* code backup

* code backup

* remove unnecessary file

* fix various bugs

* remove unnecessary changes

* remove unnecessary submodule

* remove unnecessary lines

* fix algorithm

* formatting

* refine onnx backend unit tests

* pin numpy version

* fix build issue

* fixed a filename to be copied

* add the onnx dependency in docker image

* ensure results are copied back correctly

* specify onnx version

* update excluded tests

* remove unnecessary log info

* turn on more unit tests

* restrict onnx backend test to python 3.x

* clang format

* refine retrieving the input parameters

* clang format

* fix program input parameter names

* clang format

* avoid running onnx test in python 2.x

* fix cppcheck error

* fix python2.7 backend unit tests error

* clang format

* resolve the issue of ensure data copy to be completed

* clang format

* fix review comments

* fix onnx backend unit test error

* another change to make onnx backend test pass

* clang format

* fix onnx backend test error

* clang format

* disable onnx backend test to try

* build try

* update Dockerfile to try onnx backend test

* remove unnecessary code

* fix a bug in copying program

* clang format

* update dockerfile to include onnx

* fix review comments

* add the pytest module to the container

* exclude real model to avoid to be downloaded

* resolve the sync device for data copy from gpu to cpu

* clang format

* fix review comments

* clang format

* move sync_device after memory_coloring
Co-authored-by: Khalique <15948690+kahmed10@users.noreply.github.com>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>
Co-authored-by: Paul Fultz II <pfultz2@yahoo.com>

* Add reduce dims

* Formatting

* Reduce dims on the gpu

* Formatting

* Fix tidy issues

* Convert to assert

* Reduce dims for contiguous

* Formatting

* Remove move

* Fix arguments used

* Formatting

* Fix warnings

* Formatting
Co-authored-by: Shucai Xiao <shucai.xiao@amd.com>

* code backup

* clang format

* fix compiling errors

* clang format

* rename a few files

* rename a few files

* fix variable bugs

* clang format

* add an operator to shift input sequences

* clang format

* fixed a bug

* clang format

* fixed a bug

* clang format

* code backup

* clang format

* code backup

* clang format

* code backup

* clang format

* refine code related lstm operator optimization

* clang format

* fix various bugs

* clang format

* fixed a bug in rewrite_lstm

* clang format

* fixed another bug

* refine two operator names

* clang format

* refine file names

* fix cppcheck error

* clang format

* fix cppcheck error

* clang format

* fix cppcheck error

* fixed review comments

* clang format

* add unit tests

* clang format

* add unit tests

* clang format

* refine unit tests for better coverage

* clang format

* fixed a bug

* fix cppcheck error

* fix review comments

* clang format

* rename two operators according to review comments

* clang format

* fix review comments

* clang format

* fix review comments

* clang format

* fix review comments

* fix a cppcheck error

* clang format

* fix review comments

* clang format
Co-authored-by: Shucai Xiao <scxiao@prj47-rack-99.local.lan>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* fix pad calc

* bert tf passes correctness

* formatting

* add test

* formatting

* remove comment

* add inline

* formatting

* fix order for literal

* formatting

* add test for gelu

* formatting

* added add_gelu fusion

* add files

* formatting

* remove layernorm opt

* revert reduce file

* add gelu_fn and tests

* formatting

* fix matcher, remove extra tests

* formatting

* fix matcher

* add used_once

* formatting

* start on new gelu

* formatting

* add matchers in fuse_ops

* formatting

* add dce to fix add_gelu

* add simplify_rsqrt and test

* formatting

* debugging value for matcher

* formatting

* add more to matchers

* formatting

* fix errors

* remove onnx gen

* add any_arg, change matchers to use either_arg

* formatting

* formatting

* add used_once

* formatting
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* add recip gpu and tests

* formatting

* remove to_hip_type
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* Upgrade compiler to hcc

* Add missing perl package for hcc

* Update hip and rocblas version

* Add flag to speedup build

* Remove symlink

* Add no-gpu-rdc to linking

* Add missing define

* Disable no unused argument

* Suppress warning

* Add definition to all targets

* Fix compiler flags

* Refer to rocblas

* add prelu operator

* clang format

* add prelu to gpu lowering

* add unit tests for the PRelu operator

* clang format

* add missing onnx file for PRelu operator

* update unit tests for prelu operator

* clang format
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>
Co-authored-by: Paul Fultz II <pfultz2@yahoo.com>

* Add initial api

* Formatting

* Add more api

* Formatting

* add more operators (asinh, acosh, atanh, MatMulInteger, ConvInteger)

* clang format

* add unit tests for new operators

* clang format

* Add auto api generation

* Formatting

* Fix some compilation errors

* Change handle struct

* Formatting

* Fix reamining compilation errors

* Formatting

* Simplify using ctype

* Formatting

* Initial c++ generation

* Formatting

* Add C++header

* Formatting

* Add test

* Formatting

* Add initial tests

* Formatting

* Try to fix formatting

* Cleanup formatting

* Formatting

* Fix constructors on the same line

* Fix tests

* Formatting

* Fix tidy issues

* Fix tidy issues

* Fix naming issue

* Add onnx API to parse buffer

* Formatting

* Add arguments api

* Formatting

* Fix verify parameters

* Fix cppcheck issues

* Formatting

* Add method to get output shapes and bytes

* Formatting

* Try formatting

* Formatting

* Improve the test coverage

* Formatting

* Add print method

* Formatting

* Fix cppcheck issue

* Fix package dependency

* Add nolint

* Try fix formatting

* Formatting

* formatting

* formatting

* Fix formatting
Co-authored-by: Paul Fultz II <pfultz2@yahoo.com>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>
Co-authored-by: kahmed10 <15948690+kahmed10@users.noreply.github.com>

* initial testing

* add shape op

* formatting

* add env variable for batch sizes

* formatting

* progress on driver

* progress on driver

* cleanup

* cleanup

* add and modified prev tests

* formatting

* remove comment

* add shape op test

* formatting

* manually insert shape op in test

* formatting

* create options struct for parsers

* formatting

* Add documentation for python

* Fix c++ documentaion

* add documentation to parser

* formatting

* add argmin and tests

* fix doc and definitions

* formatting

* revert test functions

* formatting

* cpu impl of conv_transpose

* more work on conv_transpose

* rename files, added extratests

* formatting

* add more tests

* formatting

* changes

* fix tests

* fix tidy

* formatting

* fixed function parameter

* fix function parameter

* add cpu ops test

* formatting
Co-authored-by: Paul Fultz II <pfultz2@yahoo.com>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* add reduce operators as required by onnxruntime

* clang format

* remove a test since it can cause overflow

* resolve cppcheck error

* clang format

* fix cppcheck error

* clang format

* Add compiler options

* Add copy operators

* Formatting

* Use run_passes in tests

* Formatting

* Use run_pass in schedule test

* Formatting

* Add compile_options to get_passes in target

* Formatting

* Offload copy option

* Formatting

* Copy using pinned memory

* Formatting

* Improve performance of gpu copying

* Formatting

* Dont copy

* Formatting

* Always make an extra copy

* Formatting

* Remove unused write op

* Add missing include

* Remove copy_to_gpu function in python api

* Make offload copy disabled by default on C++

* Formatting

* Fix tidy issues

* Formatting

* Fix namespace

* Fix python tests

* Turn clang format off since its broken

* Fix compile error on gcc 5

* Remove commented code

* Add functions to do multi-index for local strides as well

* Formatting

* Use same multi-index path for block_reduce

* Formatting

* Use multi-index calc in reduce

* Formatting

* Fix warning

* Fix compiler warning

* Disable some tidy checks

* Fix bug in bert accuraccy

* Formatting

* add another test

* Fix add and overflow

* Formatting

* Fix bug in shape_for_each

* Use front instead of iterator

* Use result.front()

* Split add_unary files

* Formatting

* Fix incorrect last index

* Remove comment

* Inline function

* Fix carry check

* Fix metadata errors

* Formatting

* Reflow

* Reflow

* add two operators ceil and floor

* clang format

* add unit test for the ceil and floor operators

* remove unintended code

* first version of refactoring reduce operators.

* clang format

* refactor the gpu implemantation of the reduce_mean operator

* clang format

* refactor gpu implementation of the resuce_sum operator

* fix cpp check error

* fix cppcheck error

* fix cppcheck error

* fix review comments

* clang format

* fix a jenkin error

* fixed review comments

* clang format

* fix review comments

* clang format

* fix review comments

* clang format

* add implemenation of reduce_min and reduce_max

* clang format

* add unit test for reduce_min/max operator

* clang format

* add more unit tests

* clang format

* fix review comments

* Remove extra copy in gemm

* combine rocblas gemm call

* clang format

* fix a bug in calling rocblas function

* clang format'

* backup of temporary changes

* clang format

* unify the gemm call to avoid multiple gpu implemantation

* clang format

* remove unnecessary code

* backup temp changes

* clang format

* fix cppcheck error

* code backup

* clang format

* remove unnecessary synchronization function

* clang format

* fix bugs

* clang format

* more optimization related to gemm

* clang format

* code cleanup

* implementation that can achieves better performance

* clang format

* temp changes to try performance

* clang format

* revert to previous commits

* fixed review comments

* clang format

* fix review comments

* add tests, fix bug in ternary op

* formatting

* uncomment fusion

* first version of refactoring reduce operators.

* clang format

* refactor the gpu implemantation of the reduce_mean operator

* clang format

* refactor gpu implementation of the resuce_sum operator

* fix cpp check error

* fix cppcheck error

* fix cppcheck error

* fix review comments

* clang format

* fix a jenkin error

* fixed review comments

* clang format

* fix review comments

* clang format

* fix review comments

* clang format