GitLab

* Remove version from check_context and bump program version

* [MLIR] add dot offloads with manual tuning support
* This commit adds dot + pointwise fusion support
along with manual tuning using rocMLIR.

Co-authored-by: Chris Austen <causten@users.noreply.github.com>

Useful to get more insight into Onnxruntime. Allows us to reuse the accuracy checker code while also allowing us to capture Execution Provider output with the --ort_run and --ort_logging flags

Also added the --target flag as well to allow us to force using either a specific target for the accuracy checking. Originally this was defaulting to the GPU. This now allows us to use ref, fpga, etc to quickly change targets.

prevent dynamically loading the target library that is not compiled with the same version of MIGraphX core lib.

Refactor to have select_module use output parameters
Disable select_module verify tests on cpu

Recent changes #1608 removed migraphx_all_target lib from driver and that led to missing compile time definitions.
Missing compile definitions led to change of default target in driver.

Fixes #1595

This is the original testcase that sparked the error with missing proper const
folding. Pushing changes up to this branch and closing out the PR #1622

* use GITHUB_STATE

* [MLIR] Adds a runtime switch to trigger MLIR

* Update github action for onnxruntime sync to open autoPR

fallback to msgpack for older msgpack versions

* NMS improvements

resolve half info messages while building 

github action file change

Enable multiple orgs for testing

Add additional documentation to explain the passes.

Creates the select_module operator that selects one of the submodules passed to it to run based on the submodule parameters.  The submodule is selected by having the exact same static shapes for the arguments to select_module as the parameters in the submodule

* Add a Changelog to the repository

Enable python 3.10 bindings

Avoids double global loads.  Strided loops are unrolled which lets store results in array which compiler will use registers for since the index access is constant.   Updated to handle large reductions so which results with a better stable diffusion result

* deprecate HCC

* Add driver flag "--exhaustive-tune" to enable tuning, add support for the same in C/C++ and python API

Add dynamic shape support to slice operator.

First draft of this feature doesn't support ops slicing non-fixed, dynamic axes. Resulting shape in such cases is not guaranteed.* Also, onnx parsing doesn't support any arguments other than "axes".

Use requirements-dev.txt instead of the other instances