GitLab

Signed-off-by: Artur Wojcik <artur.wojcik@amd.com>

Signed-off-by: Artur Wojcik <artur.wojcik@amd.com>

* Add initial tuning support

* Format

* Add extra param

* Format

* Use exauhstive flag

* Format

* Set expected shapes

* Format

* Format

* Fix missing symbol

* Format

* Add missing license header

* Format

* Update src/targets/gpu/include/migraphx/gpu/mlir.hpp

* Updates to add_embed_library

* Fix warnings for extern arrays

* Improve performance of pointwise/reduction kernels when using NHWC layouts

* Format

* Add nhwc test

* Format

* Remove inline namespace

* Add reduce test

The --py output can be loaded back in the driver. This will embed the migraphx interperter so we can execute the python directly. There is a migraphx_py library which will dynamically load the version of the library for python version is available on the system.

* rename function 'near' to 'make_near'

* try disabling vega10 machine

* Remove rocblas_gemm_flags_pack_int8x4 in rocblas v3.0 or higher (#2002)

* Update src/targets/gpu/gemm_impl.cpp

* Add dynamic input to prefix_scan_op

* Added a shape test.  The op should return the same dynamic shape as the input.

* add 2d shape test for prefix_scan

add JetBrains intermediates to ignore

Update ROCmSoftwarePlatform/half from 5.4.2 to 5.6.0

* Cancel in progress performance runs if a new commit has been pushed. Concurrency only cancels any pending runs.

Throwing on these calls catches dynamic shape errors earlier rather than having to backpedal from a bad call

onnxruntime is rolling a version with current versions 3.26

This adds a format.py file which will check or apply formatting using git clang-format or yapf.

Fixes #1957

Clamping was removed in #1853.

Turns out clamping as necessary to handle overflow/underflow cases. during downcasting, if it overflowed then without clamping it returned infinity.

Fixes #1746

BatchNorm only has x as the runtime input parameter for the following equation. All the other parameters are compile-time constants and related operations can be const-folded before quantizing to fp16 to preserve precision.