Skip to content

ROCm-Apex v0.2

Compare
Choose a tag to compare
@lcskrishna lcskrishna released this 29 May 21:39
· 484 commits to master since this release
aea81c0

ROCm Apex v0.2 Release Notes:

  • BFloat16 Mixed Precision Training support with single GPU.
    • Introduced new Optimization levels - O4 and O5 for BFloat16 training.
    • BFloat16 support for FusedOptimizers
    • BFloat16 support for FusedLayerNorm.
  • Performance Improvements.
  • Test Infrastructure and CI setup on ROCm.