GitLab

* remove forced FP32 modules

* correct no_shard-contexts' positions

* [test] refactored with the new rerun decorator

* polish test case

* adapt post grad hooks for not-shard parameters
* adapt optimizer for not-shard parameters
* offload gradients for not-replicated parameters

* adapt model weight initialization for methods in Pytorch nn.init

* support existing sharded and unsharded parameters in zero

* add unitest for moe-zero model init

* polish moe gradient handler