Merge pull request !4019 from yao_yf/add_layout_in_stra_ckpt

Merge pull request !3555 from fary86/support_dynamic_min_max_shape

Merge pull request !3980 from Chong/wd

Merge pull request !3789 from riemann_penn/eager_mode_sparse

Merge pull request !3950 from ZPaC/ci-add-ps-cases

Merge pull request !3966 from lichen/add_dropout_distributed_op

Merge pull request !3720 from thlinh/dev_July29_remove_redundant_minmax

Merge pull request !3928 from yihuaijie/dev

Merge pull request !3715 from BowenK/new_pattern

Merge pull request !3370 from xychow/fix-context-dup

Merge pull request !3900 from ZPaC/master-delete-hard-code-in-pull-node

Merge pull request !3858 from ZPaC/master-fix-ps-sgd

Merge pull request !3103 from yihuaijie/dev

Merge pull request !3820 from zhangbuxue/support_dtype_and_shape_as_attr_in_graph_mode

Merge pull request !3473 from vlne-v1/change_base_class_of_parameter

if funcgraph or metafuncgraph call it recursively, then anf_node should be used as tracking_id to discriminate the first occurcance and the
recursive occurance.
add anf_node to PrimitiveAbstractClosure hash() to reduce cost of GetEvaluatorFor().

ignore the tracking_id to make cse work.

Merge pull request !3859 from yao_yf/embeddinglookup_parallel_ops_host_device

Merge pull request !3744 from Xiaoda/14-add-new-graph-operations-in-auto-parallel

Merge pull request !3537 from riemann_penn/add_sparse_operator

Merge pull request !3636 from Kang/opt

Merge pull request !3657 from yangzhenzhang/add-stridedslice-op

Merge pull request !3405 from gziyan/fix_optimizer_parallel

Update threshold for rounding when checking expected value in input tensor node