main
[pre-commit.ci] pre-commit autoupdate (#5572)
[shardformer] added development protocol for standardization (#4149)
[upgrade]Upgrade transformers (#6320)
[shardformer] support bias_gelu_jit_fused for models (#5647)
[Device]Support npu (#6159)
[moe] add parallel strategy for shared_expert && fix test for deepseek (#6063)
[shardformer] support pipeline for deepseek v3 and optimize lora save (#6188)
[misc] update pre-commit and run all files (#4752)
[Distributed RLHF] Integration of PP (#6257)
[pre-commit.ci] auto fixes from pre-commit.com hooks