* add gemm_int8 arm64 version without openmp

* add gemm_int8 arm64 version with openmp

* update concat and split kernel and related files in FPGA v2(v3) track

* update

* update

* update kernel and related files in FPGA v2 track

* update

* update

* update kernel and related files for static quantization in FPGA v2 track

* update

* update feed and fetch kernel in FPGA v2 track

* update io file

* update feed fetch and softmax kernel in FPGA v2 track

* update proposal kernel and other kernels in FPGA v2 track

* update fetch and softmax kernel in fpga v2 track

* update v1 and v3 kernel file in FPGA track

* add auto debug tools

* fix style

* update concat and split kernel and related files in FPGA v2(v3) track

* update

* update

* update kernel and related files in FPGA v2 track

* update

* update

* update kernel and related files for static quantization in FPGA v2 track

* update

* update feed and fetch kernel in FPGA v2 track

* update io file

* update feed fetch and softmax kernel in FPGA v2 track

* update proposal kernel and other kernels in FPGA v2 track

* update fetch and softmax kernel in fpga v2 track

* marker1-anchor update

* concat-fetch bug

* update concat and split kernel and related files in FPGA v2(v3) track

* update

* update

* update kernel and related files in FPGA v2 track

* update

* update

* update kernel and related files for static quantization in FPGA v2 track

* update

* update feed and fetch kernel in FPGA v2 track

* update io file

* update feed fetch and softmax kernel in FPGA v2 track

* update proposal kernel and other kernels in FPGA v2 track

* add reshape2, transpose2, split GPU operator

* fix reshape2op && delete log

* [TEST] add gesture model for test

1. enable conv mps with texture bias. 2. add scale mps. 3. remove duplicate computation in fetch kernel. 4. fix typo. (#1665)

* 1.enable both tensor and texture as bias of conv_add_relu. 2. add conv_relu fusion op.

* winograd could be used when. out channel is less than or equal to 16

1.update CPUContext to probe CPU info 2.improve the performance of SlidingwindowConv3x3s1 and SlidingwindowConv3x3s2 (#1655)

* fix opencl no persistable bias in fusion conv op

* delete log

* fix cpu conv_add and conv_add_relu

* fix cpu conv_add and conv_add_relu

1. add optimize parameter to load function. 2. add transpose operation if the input of fetch op is not [0, 2, 3, 1]. (#1651)

* develop nearestinterp op  fix bug in element add close #1647

* remove log
remove log

1. reshape output transpose should be 0, 1, 2, 3. 2.mps conv should be disabled when groups is larger than 1. 3. slice should re-tranpose its input if needed. (#1646)

* fix memcpy size in opencl fetch kernel

* add opencl Scale op

1.add groups to conv3x3. 2.add original size to tensor and tenture and adjust concat layer to handle tensor truncation logic. (#1643)