fast rcnn cudnn 5.0
2017.01.08 15:47
手动修改文件,参考了卜居大神的博客 http://blog.csdn.net/kkk584520/article/details/51163564方案1简单方便,但是当我们编译的是其他人修改过得源码,可能就会出错。方案2 步骤如下:
1. 用最新caffe源码的以下文件替换掉faster rcnn 的对应文件
include/caffe/layers/cudnn_relu_layer.hpp, src/caffe/layers/cudnn_relu_layer.cpp, src/caffe/layers/cudnn_relu_layer.cu
include/caffe/layers/cudnn_sigmoid_layer.hpp, src/caffe/layers/cudnn_sigmoid_layer.cpp, src/caffe/layers/cudnn_sigmoid_layer.cu
include/caffe/layers/cudnn_tanh_layer.hpp, src/caffe/layers/cudnn_tanh_layer.cpp, src/caffe/layers/cudnn_tanh_layer.cu
2. 用caffe源码中的这个文件替换掉faster rcnn 对应文件
include/caffe/util/cudnn.hpp
3. 将 faster rcnn 中的 src/caffe/layers/cudnn_conv_layer.cu 文件中的所有
cudnnConvolutionBackwardData_v3 函数名替换为 cudnnConvolutionBackwardData
cudnnConvolutionBackwardFilter_v3函数名替换为 cudnnConvolutionBackwardFilter
댓글 0
번호 | 제목 | 글쓴이 | 날짜 | 조회 수 |
---|---|---|---|---|
274 | slide note 삭제 슬라이드 노트 삭제 | WHRIA | 2018.09.12 | 4439 |
273 | Deep learning 기반 DEMO | WHRIA | 2018.09.16 | 7122 |
272 | register | WHRIA | 2018.09.21 | 1937 |
271 | ambient-light-sensor | WHRIA | 2018.09.29 | 283 |
270 | Intro | WHRIA | 2018.10.01 | 4088 |
269 | faster rcnn cuda 10 | WHRIA | 2018.10.06 | 2109 |
268 | delayed network drive connect | WHRIA | 2018.10.07 | 152 |
267 | mixed precision training [1] | WHRIA | 2018.11.06 | 1781 |
266 | excel 에서 소수점 표시 함수 | WHRIA | 2018.11.08 | 202 |
265 | ac-68u dd-wrt 에서 정펌으로 복구 | WHRIA | 2018.11.22 | 548 |
264 | fatal error: turbojpeg.h: No such file or directory | WHRIA | 2018.11.24 | 646 |
263 | CNN Models | WHRIA | 2018.12.23 | 353 |
262 | imbalanced dataset | WHRIA | 2018.12.26 | 45 |
261 | center % crop | WHRIA | 2018.12.27 | 72 |
260 | Swap file add | WHRIA | 2019.01.02 | 2132 |