* fix pooling3x3s2 max. test=develop

* fix format. test=devleop

* fix format. test=develop

* remove paddle mobile old project.  never say good bye

* test=develop

* add device info for 865. test=develop

* [LITE][BM] support multiclass_nms2 and fix some issues, test=develop

* create

* [LITE][BM] fix input shape order changed issue,test=develop

* [Bugfix][OpenCL][Core] fix opencl multi-run result error when using memory_optimize_pass. test=develop

* Add test for int16 quantized model, test=develop

* [ASCEND] fix build for Ubuntu18.04 and GCC7.3, test=develop

* [ASCEND] address review comments, test=develop

* add clear_blocks interface for flatbuffers view, test=develop

* fix a bug, test=develop

* enhanced the error messages, test=develop

* add reduce op on arm. test=develop

* fix format. test=develop

* fix acccording to comments. test=develop

* fix bilinear_resize result not equal with fluid. test=develop

* fix cv build error. test=develop


* fix format. test=develop

* fix conv_3x3s1_dw v7-compute nan. test=develop

* fix compute. tets=develop

* set sgemm basic_test is false. test=develop

* [X86] add box_coder kernel and fix compile on MAC, test=develop

* [DOC] remove debug info, test=develop

* [X86] add new op of density_prior_box, test=develop

* add scatter op on arm. test=develop

* fix cmake error. test=develop

* test=develop

[PROFILE] Add ENV var controls whether write output tensor of each op to files; Rename output tensor name when mem_reuse pass enabled by default etc. (#4348)

*  Add ENV var controls whether write output tensor of each op to files;
*  Rename output tensor name when mem_reuse pass enabled by default etc.

* [DOC] fix doc of support models tutorials and arch, test=develop, test=document_fix

* [DOC] update architecture, test=develop, test=document_fix

* [PROFILE] Write output tensor to file for each OP when precision profiler enabled. test=develop

* create output tensor files dir. test=develop

[LITE][XPU] 1. Add sequence_unpad kernel for XPU; 2. Bugfix in sequence_unpad kernel for x86, as InferShapeImpl() is now empty in lite/operators/sequence_unpad_op.cc; 3. Refine TargetWrapperXPU; (#4237)

* [cherry-pick] not build naive_buffer when build light framework, test=develop (#4251)

* fix opt fatal errors, test=develop (#4293)

* fix cmake dependencies, test=develop

* [Bugfix][OpenCL] fix depthwise_conv2d_3x3 with dilation > 1. test=develop