* [NPU] subgraph's precision register to kAny

* add mean op, test=develop

* split forward and backward, test=develop

* add lookup_dequant_op

* enhance gather, lookup_table arm kernel uts

* enhance beam_search, beam_search_decoder, increment rigster

* get the precision from the real tensor or tensor_array of base_scope in arena/framwork

* register assign and assign_value to kAny

* scale support int32 input

* [NPU] support fill_constant and fill_constant_batch_size_like op bridge and uts

* add sequence_conv op and arm kernel

* add test, test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* modify code style. test=develop

* fix ut, test=develop

* delete unused code, test=develop

* Support lac model, fix bug for argmax lod_reset, test=develop

* Add reduce sum op test
test=develop

* Optimize the InferShape of several operators.
test=develop

* Remove the new function, resize and CheckPositive in DDim.
test=develop

* Fix a bug in fc_op's InferShape.
test=develop

* refactor 5x5s1 dw conv armv8, test=develop

* [ARM] refactor depthwise conv 5x5s1, and support relu6, leakey relu, test=develop

* [ARM] sgemv support fuse relu6 and leakey relu,test=develop

* [ARM] reduce some conv ut case, test=develop

* [ARM] fix 5x5dw conv pick kernel bug, test=develop

* fix code style, test=develop

* [ARM] fix sgemv fuse relu6 bug, test=develop

* [ARM] fix fp32 5x5s1 dw bug, test=develop

* [ARM] fix fp32 5x5 dw conv pick kernel bug, test=develop

* [NPU] reshape x,y,out node in elementwise ops

add gemm + relu6/leakyrelu fusion

- enhance interp InferShape

* Support bitman backend

[LITE][NPU] Add instance_norm op bridge and unit test, refine the registration of op bridges (#2747)

* [NPU] enhance unittest for shuffle_channel, unsqueeze, pool

test=develop

test=develop

test=develop

X86 and cuda compile simutaneously cmake ..  -DCMAKE_BUILD_TYPE=RelWithDebInfo  -DWITH_MKL=ON           -DLITE_WITH_CUDA=ON           -DWITH_MKLDNN=OFF           -DLITE_WITH_X86=ON           -DLITE_WITH_PROFILE=OFF          -DWITH_LITE=OFF           -DLITE_WITH_LIGHT_WEIGHT_FRAMEWORK=OFF           -DWITH_PYTHON=OFF           -DWITH_TESTING=ON           -DLITE_WITH_ARM=OFF           -DLITE_ON_TINY_PUBLISH=OFF           -DCUDNN_ROOT=/usr/local/cudnn/           -DLITE_BUILD_EXTRA=ON (#2708)

x86 and cuda compile simutaneously

test=develop