Skip to content

MIGraphX v0.4
Compare
Choose a tag to compare
@mvermeulen mvermeulen released this 11 Oct 19:20
· 1844 commits to master since this release
0.4
51bd00b

Notes

Updated release

  • Additional operators support for new models and features
  • Quantization support for fp16 and int8
  • Support for NLP models, particularly BERT with both Tensorflow and ONNX examples

See https://github.com/ROCmSoftwarePlatform/AMDMIGraphX/wiki/Getting-started:-using-the-new-features-of-MIGraphX-0.4 for more getting started information

Assets 6
Loading