[Common, PyTorch] Improve mHC to match DeepSeek's implementation#2953
Draft
kainzhong wants to merge 10 commits intoNVIDIA:mainfrom
Draft
[Common, PyTorch] Improve mHC to match DeepSeek's implementation#2953kainzhong wants to merge 10 commits intoNVIDIA:mainfrom
kainzhong wants to merge 10 commits intoNVIDIA:mainfrom
![pre-commit-ci[bot]](https://avatars.githubusercontent.com/in/68672?v=4&size=32){kind=avatar}