GitLab

When multiplying either the input or output across the K dimensions then the multiple can be applied to the constant which can then be folded with propagate_const.

Improves the constant propagation for bert models. Larger batch size no longer use as large of constants.  Also improves the speed of model compilation 

This will replace conv(x+a, w) with conv(x, w) + conv(a, w) where a is a constant so conv(a, w) can be replaced with a constant.

Using add_instruction for the neg op was causing issues on replace_instruction. Changed to use insert_instruction.  Tests and added a new one that is failing without the change.

* Fuse layernorm with different patterns
* Only match when using the last axis
Co-authored-by: kahmed10 <15948690+kahmed10@users.noreply.github.com>
Co-authored-by: kahmed10 <15948690+kahmed10@users.noreply.github.com>

This PR resolves the bug addressed in #1496. 

Upgraded Dockerfiles and fixed tidy issues to make Ubuntu 20.04 and ROCm 5.3.0 the default 

Simplified algebraic operations (x*1), x*(-1), x/1, 0+x & x+0,  x-0, 0-x, 0*x, x*0, and 0/x operations

* Fix invalid program from find_splits

Change find_mul_conv to work with multibroadcast also. Checks the strides instead of the broadcast axis.

* fix slice_dim1 for case

Using not and or improves readability. The cppcheck rule will help ensure we are doing it consistently.

reshape op requires standard shape. During simplify_algebra, it inserts reshapes without checking for this requirement.

This will rewrite dot operators like X(Y + b) to XY + Xb when b is constant as we can fold the add away.
This improves handling pointwise with broadcasted operators, this helps improves const propagation.
Improve gemm fusion with a mul_add
Improve support for broadcast shapes in gemm

Updated each source file in the repo with the existing license.

Updated variable names according to #1193

Fuse layernorm and added triadd_layernorm fusion.  This is a prep performance booster

Fixes the #error when using cppcheck. This no longer suppresses cppcheck errors when including those errors. This fixes the cppcheck errors that was there already.

Update the base version of clang-format from 5.0 to 10.0

In migraphx, DepthToSpace (d2s) is implemented as reshape --> transpose --> contiguous --> reshape.

This PR adds matcher to find d2s + unary pointwise ops.

Application of the matcher moves the pointwise unary operation before the contiguous and reshape of the d2s.
So it becomes
reshape --> transpose --> unary --> contiguous --> reshape.

Motivation is that, later pointwise module would be created out of unary --> contiguous --> reshape. Codegen for this pointwise module can write out buffer such that explicit contiguous and reshape wouldn't be required.

This transformation is not always guaranteed to improve performance, since unary op will operate on non-standard shape. So, we would need some tuning mechanism to make decision.

#905 pending PR for binary operations.

Adds a pass to fuse pointwise operators into one "pointwsie" op that has a submodule which does the calculation.

* rename broadcast and multibroadcast output_lens attribute to out_lens attribute, and change tests and source code to reflect the same

* change the reshape attribute from dims to out_lens

* change transpose attribute's name from dims to perm to reflect better meaning

* use permutation instead of perm for transpose

clang formaating

* use dims instead of out_lens for reshape

clang formatting

* Enable libstdc++ debug mode

* Add is_end function

* Compare addresses in a map or set

* Formatting

* Check end

* Fix comparision of instruction_ref

* Formatting

* Some more iterator fixes

* Formatting

* Fix assert

* Fix invalid iterators

* Fix debug print in program

* Remove debug flag for now

* Set correct bool type
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* add the operator step

* clang formatJ

* add unit tests

* clang format

* add more unit test for step op

* clang format

* add more unit tests

* clang format

* fix review comments

* clang format

* rename two unit tests
Co-authored-by: Paul Fultz II <pfultz2@yahoo.com>

* backup implementation of resize enhancement

* clang format

* code backup for the resize

* clang format

* fix build error for resize operator

* clang format

* tmp code backup

* clang format

* remove changes in parse_resize

* remove unnecessary changes

* clang format

* add unit test for the bug

* clang format

* remove print code

* remove a semi-colon

* clang format

* fix a tidy error

* fix review comments

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

* code backup

* clang format

* add a matcher related to the special resize case for optimization

* clang format

* code backup

* clang format

* code backup

* remove unnecessary code

* add optimization for the where op

* clang format

* fix cppcheck error

* add a unit test for optimize resize

* clang format

* remove unnecessary header include

* code backup

* clang format

* add unit tests for optimizing resize

* clang format

* add more unit test for optimizing where op

* clang format

* remove unnecessary code

* add one more optimzation to remove contiguous

* clang format

* add a pointwise requirement

* clang format

* fix cppcheck error

* add one more unit test

* fixed a bug

* clang format

* remove unnecessary code

* clang format

* fix a build error

* fix review comments

* clang format

* fix a review comments

* clang format

* code refinement

* clang format

* refine more code

* refine more code

* fix a bug related to reshape_cont optimization

* clang format

* fix a review comment

* removed an unnecessary comment

* refine code according to comments

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

* fix issue#727

* clang format

* refine unit tests

* fix cppcheck error

* fix review comments

* refine a unit test to cover more code changes

* fix cppcheck error

* remove unnecessary include file

* fix review comments

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

* Load op when serializing

* Formatting

* Add missing clip field

* Use make_op almost everywhere

* Formatting

* More make ops for rnns

* Get rid of spaces

* Formatting

* Remove operators headers

* Formatting

* Remove unused op headers

* Increase line threshold

* code backup

* clang format

* change corresponding tool files

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

* Add pointwise attribute

* Formatting

* Fix compilation

* Remove unused variable

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

* Add initial optimization when using a mul over a sliced convolution

* Formatting

* Add more tests

* Formatting

* Convert to an assert

* Check if used once

* Formatting

* Add test with horiz fusion

* Formatting

* Optimize nested slice

* Formatting

* Fix test

* Add const refs

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

* code backup

* clang format

* refine the algorithm to support more scenarios

* clang format

* fix review comments

* clang format

* add one more unit tests to have more code change coverage
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* fix pad calc

* Add decompose pass

* Add decompose test

* Formatting

* bert tf passes correctness

* formatting

* Add remap

* Formatting

* add test

* formatting

* remove comment

* Add compute method for dot

* Formatting

* add inline

* Add finder for horizontal fusion

* Formatting

* Formatting

* Reuse predicate

* formatting

* fix order for literal

* formatting

* add test for gelu

* formatting

* added add_gelu fusion

* Add gemm fusions

* Formatting

* add files

* formatting

* test no mul_add

* formatting

* progress on div

* formatting

* continue work on pass

* remove layernorm opt

* revert reduce file

* Add some fixes for convolution

* Formatting

* Fix shape tests

* Formatting

* Reuse axis equal

* Add initial split fusion

* Formatting

* Update offset

* Workaround outputs that cant accept nonstandard shapes

* Formatting

* Add check for split concat

* Formatting

* Add missing headers

* Formatting

* Add tests

* Formatting

* add optimization for bert

* code backup for bert optimization

* continue testing

* formatting

* fix matcher

* formatting

* add gelu_fn and tests

* formatting

* fix matcher, remove extra tests

* formatting

* fix matcher

* add missing files

* add find_layernorm

* add add_transpose to cmake file

* code backup for the contigous fusion

* refine ops fusion

* clang format

* fixed bug in previous optimization

* clang format

* add more optimization

* remove unnecessary code

* refinement of the fustion code

* clang format

* fixed a bug

* 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

* clang format

* formatting

* add used_once

* formatting

* code cleanup

* clang format

* fixed a bug

* remove unnecessary code

* refine comments

* optimize bert to remove more contiguous

* clang format

* remove unnecessary code

* add unit tests for bert optimization

* clang format

* fix review comments

* clang format

* refine a fusion of reshape and slice

* clang format

* fix cppcheck error

* fix review comments

* add the fusion of slice and transpose

* clang format

* add another optimization to fuse slice and transpose

* clang format

* fix review comments

* clang format

* fix review comments

* clang format

* fix review comments
Co-authored-by: Khalique <15948690+kahmed10@users.noreply.github.com>
Co-authored-by: Paul <pfultz2@yahoo.com>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>
Co-authored-by: Shucai Xiao <scxiao@prj47-rack-99.local.lan>

* Fuse operators that partially match in a concat operator

* Formatting

* Remove unused matcher

* Properly calculate the output_lens

* Formatting

* Formatting

* Formatting

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

* Skip fusing group convolutions

* Formatting

* Fix ICE on gcc 5

* Formatting

* Fix gcc check

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

* Ensure the slices are in order before removing them in concat

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

* Add decompose pass

* Add decompose test

* Formatting

* Add remap

* Formatting

* Add compute method for dot

* Formatting

* Add finder for horizontal fusion

* Formatting

* Formatting

* Reuse predicate

* Add gemm fusions

* Formatting

* Add some fixes for convolution

* Formatting

* Fix shape tests

* Formatting

* Reuse axis equal

* Add initial split fusion

* Formatting

* Update offset

* Workaround outputs that cant accept nonstandard shapes

* Formatting

* Add check for split concat

* Formatting

* Add missing headers

* Formatting

* Add tests

* Formatting

* Add more testing

* Formatting

* Fix when there is duplicate splits in inputs

* Formatting

* Fix mismatch iterators

* Add tests for dot fusions

* Formatting

* Add test for convolution

* Formatting

* Fix tidy issues

* Add more tests

* Formatting

* Ignore build directory for codecov

* Add test for groups

* Formatting

* Add more tests for groups

* Formatting

* Add test for missing end slice

* Add newline

* Remove unused function

* Add support for when beta is not 1

* Formatting

* Add test for scalar

* Add one more scalar test
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* add simplify_rsqrt and test

* formatting

* add used_once check

* move used_once

* formatting

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

* Add test with odd stride

* Use same logic as convolution for stride calculation

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