* add use_hierarchical_allreduce

* support hierarchical allreduce for more cases

* remove cudaDeviceContext

* remove more template

* fix rocm compile

* remove alias name CUDADeviceContext

* fix compile

* fix tests

* revert changes

* add nproc_per_node for DistributedFusedLamb

* fix nproc_per_node communicator bug

* fix ring_id = 1 init bug

* fix ci

* fix test_parallel_executor_mnist.py

* add use_master_acc_grad

* add ut

* add gradient merge for DistributedFusedLamb

* use master acc gradient

* fix CI ut

* polish

* remove math_function_impl.h change

* fix test_update_loss_scaling_op.py

* try to fix XPU/NPU CI

* add gm ut

* add Output(Step) to distributed fused lamb op

* add _set_step

* clean distribution_helper, index_impl, aligned_vector code in fluid

* fix conflicts

* vectorize lamb kernel

* remove flags, add ut

* remove useless codes

* refine code, add param order

* add multi tensor apply l2 norm

* add multi_tensor_apply code

* make sizeof(TensorMeta) smalller

* move code to distributed_fused_lamb_op.cu

* remove useless FLAGS

* rename pten dir to phi

* rename namespace to phi

* rename infrt pten dir to phi

* resolve conflict

* rename pten to phi in cmake

* revert all infrt change

* change needed files

* fix infrt failed

* fix inference failed

* add DistributedFusedLamb op

* polish code

* fix compile error

* compatible with pten changement

* fix rocm compile error

* improve converage

* update upstream/develop

* fix cast_with_ptr.h

* add FLAGS_distributed_lamb_divide_nranks_when_allreduce=1

* fix clip before allreduce

* add use_master_param_norm

* code polish

* fix bug

* fix ROCM ci