Releases: ROCm/AMDMIGraphX
Releases · ROCm/AMDMIGraphX
rocm-5.0.2
No changes since rocm-5.0.0
rocm-5.0.1
No changes since rocm-5.0.0
ROCm Rel 5.0.0
Support for ROCm Release 5.0.0
- Added ONNX Operators (Topk, Where, Loop, DepthToSpace, Multinominal, SpaceToDepth, Nonzero, ROIAlign, ThresholdedRelu, NonMaxSuppression (CPU), RandomNormal, RandomNormalLike, RandomUniform, and RandomUniformLike)
- ABI version bumped to 2.0
- Tool improvements (Batch size is now printed, Docker default moved to ROCm 4.5, Added support for ROCtx integration)
- Performance improvements
- Added --fp16 verify flag for reference validation
- Fix an issue that missed including the Half package as part of the install
- Added a new command to list supported ONNX Operators "migraphx-driver onnx --list"
- created macro to switch between OneDNN and ZenDNN API calls
- Improved examples and Documentation
- CI and Build improvements
ROCm Rel 4.5.2
Support for ROCm Release for 4.5.2
- Failing fusion plan workaround
- Install pcre from github since the ftp.pcre.org site is no longer available
ROCm Rel 4.5.0
ROCm Release for 4.5.0
- Moved migraphx to version 1.3
- Added support for Ubuntu 20.04
- Added support and updates for ONNX Operators scan, scatter, reverse, step, and flatten
- Improvements to build, reduced time required for model compile phase
- General improvements to Python and C++ examples along with adding Notebook and BERT SQuAD, NFNet, Unet, and 3DUnet examples
- Added Quantization support for FP16 operations
- Support running ONNXRT test suite
- Enhanced debug capabilities by logging time,
- Add support for step attribute for ONNX slice operator
- Internal Build / Test improvements
ROCm Rel 4.2.0
ROCm Release for 4.2.0
ROCm rel 3.9.0
Selu operator (#642) * code backup * clang format * support for sele operator * clang format * added an onnx unit test for selu * clang format * add more unit tests for the selu operation Co-authored-by: mvermeulen <[email protected]>
ROCm Rel 3.10.0
Selu operator (#642) * code backup * clang format * support for sele operator * clang format * added an onnx unit test for selu * clang format * add more unit tests for the selu operation Co-authored-by: mvermeulen <[email protected]>
MIGraphX v0.6
Notes
MIGraphX 0.6 is primarily a bugfix release. The release includes:
- Added a C API in addition to existing C++
- Support multiple outputs
- Additional operators including ArgMax, ArgMin, Ceil, ConvTranspose, InstanceNormalization, ReduceL1, ReduceL2, ReduceLogSum, ReduceProd, ReduceSum, ReduceSumSquare, Shape
MIGraphX v0.5
Notes
MIGraphX 0.5 is primarily a performance and bugfix release. The release includes:
- Additional operators: split, ceil, floor
- Support for additional models including NASNet-a_Large for Tensorflow
- Simplified python interface
- Performance improvements
- A driver for exercising migraphx
See https://github.com/ROCmSoftwarePlatform/AMDMIGraphX/wiki/Getting-started:-using-the-new-features-of-MIGraphX-0.5 for more getting started information