add cuda c++ doc

- 增加cuda c++ demo.
- 考虑到检测模型尾部一般是multiclass_nms，该kernel为host，如果fetch kernel为cuda的话，则会在此处插入无用的io_copy(host->cuda)，由于该原因，注释掉fetch的cuda kernel. 默认使用host的fetch kernel. 此处暗中进行的行为：每次predictor run完，都会默认把数据从cuda拷贝到cpu

* * support download bm_sdk, test=develop

* [LITE][BM] add slice op

* [LITE][BM] fix concat issue

* [LITE][BM] support reduce full ops, test=develop

* [LITE][BM] change test_resnet50 to change test_classify

* [LITE][BM] add cast op

* [LITE][BM] add reduce and other ops, test=develop

* [LITE][BM] add reduce,cast and other ops, test=develop

add cuda ci.

Organize cuda compilation products.

* add sgd op, test=develop

* test=develop

* add backend

* add kernel

* add grad kernel

* [NPU] subgraph's precision register to kAny

* [LITE][OPENCL] Change fp32 fc to fp16's. test=develop

* fix act in conv3x3opt opencl kernel. test=develop

* [LITE][OPENCL] fix opencl fc kernel. test=develop

* [LITE][OPENCL] add lws for opencl conv image kernel. test=develop

* [LITE][OPENCL] Change fp32 fc to fp16's. test=develop

* fix act in conv3x3opt opencl kernel. test=develop

* [LITE][DEMO] fix input arg. test=develop

* fix shutdownlog for std::endl. test=develop

* [LITE][OPENCL] Change fp32 fc to fp16's. test=develop

* fix act in conv3x3opt opencl kernel. test=develop

* fix clang v7 error
* change ut, test=develop
* fix conv ut, test=develop

* [Core] Fix findar for Clang
test=develop

* [CORE] Fix the missing of feed and fetch kernel when using light API to loade the optimized model
test=develop

* add mean op, test=develop

* split forward and backward, test=develop

add pad2d image kernel and ut

* feat: add dropout opencl kernel

* fix v7 clang bug. test=develop

* fix format, test=develop

* fix .a build bug, test=develop

* [NPU] revert shape check for input&output

* [NPU] reshape elt input&output

* feat: add opencl elementwise_sub op & ut

add instance norm kernel and ut

* set arm_lang default is off. test=develop

* fix resize error, test-develop

* add lookup_dequant_op

* feat: add opencl elementwise_sub op & ut

* enhance gather, lookup_table arm kernel uts

* enhance beam_search, beam_search_decoder, increment rigster

[LITE][OPENCL][Image] revert router of depthwisespl --> fix mnasnet and nanoyolo ,test=develop (#3109)

* add lrn op

* fix v7 build error, test=develop

* [Lite][OpenCL]fix depthwise_conv3*3 test. test=develop

* [Lite][OpenCL]fix depthwise_conv3*3 test, format codestyle. test=develop

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

* register assign and assign_value to kAny

* fix

* fix format. test=develop

* fix format, test=develop

* fix resize error

* fix format. test=develop

* fix format, test=develop

* fix format

* note log(4) << x_image. test=develop

* note x_image printf. test=develop

* [MLU] support resnet50 on MLU

* [MLU] support resnet50 on MLU