DeepSpeed
fix MegatronLayerPolicy to be compatible with the newest ParallelTransformerLayer
#4236
Merged

Commits
  • fix MegatronLayerPolicy to be compatible with the newest ParallelTransformerLayer
    Chen, Zhenhuan committed 3 years ago
  • Merge branch 'master' into fix-megatron-gpt
    RezaYazdaniAminabadi committed 3 years ago
Loading