* add option for cooling down

* add option for cooling down
add the option usage
add benchmark of RTX2080

* add  benchmark of rtx2060

* use Mat class for Shape description

* shape specialization constant in compute shader

* wip

* wip

* test forward_inplace, add binaryop unaryop sigmoid test

* fix arm unaryop test

* fix arm binaryop test

* make shape hint optional, cast int8 to fp32, add cast test

* wip

* follow the good and old local size setting for conv1x1

* the optimal local size rewrite

* fix build on msvc

* add permute shader for all packing layout, add permute test

* concat and slice patial shape constant, slice test

* fix slice test

* interp test

* add lrn test, test packing layout implicitly

* add eltwise test

* add normalize test

* add instancenorm test

* reorg shape constant

* simple local group size partition

* add shape constant param

* add folder property for a better looking in visual studio or other property supported IDEs

* fix condition for not found protobuf

* 1. capitalize to lowercase
2. rename visual folder 'test' to 'tests'

* Change DataReader::read()'s signature to fix warning C4267

This CL fixes lots of warning "C4267: 'initializing': conversion from
'size_t' to 'int'" in our codebase by matching DataReader::read()'s
signature to fread().

* Fix warnings C4244 and C4267 in tools/ncnnoptimize.cpp

C4244: 'initializing': conversion from 'double' to 'float',
    possible loss of data
C4267: 'initializing': conversion from 'size_t' to 'int',
    possible loss of data

* Fix warning C4244 in src\layer\selu.cpp

C4244: '=': conversion from 'double' to 'float', possible loss of data

* Fix warning C4244 in src\layer\cast.cpp

C4244: '=': conversion from 'double' to 'float', possible loss of data
C4244: 'return': conversion from 'float' to 'signed char', possible loss of data

* Fix warning C4244 in src\layer\psroipooling.cpp

C4244: 'initializing': conversion from 'double' to 'float',
    possible loss of data
C4244: 'initializing': conversion from 'double' to 'int',
    possible loss of data

* fix benchmark compilation errors

* some arm-linux toolchain file also defines ANDROID, it will break its build for missing libandroid.so

* Add mobilenetv2_yolov3 ,and remove old model

* Recovery yolov2

* Update link

* CMake improvement

* Fix bugs

* Fix typo

* Propagate vulkan dependency

* import vulkan

* add config files, now exported target cmake should be able to find packages

* Propagate no-rtti and no-exception

* Provide a option to control rtti and exception in mobile platform

* Make cmake clean

* Resolve conflicts

* Update CMake
PIE is propagated by INTERFACE_POSITION_INDEPENDENT_CODE

* Remove bad things

* initial the ncnn post training quantization tools

* clear some comments of tools

* fix the Travis ci compiler error

* Add CPU benchmark of Qualcomm Snapdragon 835

Add CPU benchmark of Qualcomm Snapdragon 835

* Add CPU benchmark of Qualcomm Snapdragon 835

* Add CPU benchmark of Qualcomm Snapdragon 835

* Add CPU, GPU benchmark of Qualcomm Snapdragon 835

Qualcomm MSM8998 Snapdragon 835 (Kyro 2.45GHz x 4 + Kyro 1.9GHz x 4 + Adreno 540)

* Update README.md

* Update README.md

* optimize the conv sgemm int8 on arm64-v8a platform

* optimize int8 arm64-v8a with sadalp ins

* merge requantize op  into latest conv layer

* merge requantize op into conv-int8 op

* update the mobilenet.param in the benchmark

* Update README.md

update Kirin970 and RK3399

* try to fix the travis build error

* add the armv7a conv3x3s1 implement without overflow,remove old codes

* fix the bug of conv3x3s2 packed int8

* new int8 implement,weight quant by perchanel,better accuracy~

* fix the bug of conv3x3s1 packed int8 neon

* add the naive c fp32 and int8 winograd F(2,3)

* add the neon intrinsic int8 winograd F(2,3)

* optimize the armv7a int8 winograd F(2,3) with neon assembly

* optimize the armv7a int8 winograd F(2,3) input transform with assembly.

* add the requantize layer and int8 relu implement.

* add graph optimize conv1x1s2 -> conv1x1s1,begin optimize int8 aarch64.

* fix int8 bugs

* add the c naive im2col with sgemm

* add aarch64 int8 winograd f23, conv3x3s2 naive implement

* add the int8 sgemm conv7x7s2 on x86/armv7a platform

* optimize the int8 sgemm by neon intrinsic and packed kernel

* optimize the int8 sgemm with packed data

* optimize the int8 sgemm with armv7a neon assembly

* add the int8 sgemm on arm64-v8a platform

* perpare to merge latest codes from master

* add the int8 param files

* In the Class Net,add the fuse_network method