Py-faster-rcnn: Check failed: error == cudaSuccess (8 vs.0) invalid device function

[wujiyang]

when I run ./tools/demo.py ,this error happens. here is the debug infomation and I am using a small net ZF:

Loaded network /home/wujiyang/py-faster-rcnn/data/faster_rcnn_models/ZF_faster_rcnn_final.caffemodel
F0319 16:42:04.865949 3385 roi_pooling_layer.cu:91] Check failed:error == cudaSuccess ( 8 vs. 0 ) invalid device function
* Check failure stack trace: *
abort(core dump)

I'm using a GPU: Nvidia Geforce GTX760 and my Makefile.config is

CUDA_ARCH := -gencode arch=compute_20, code=sm_20 \
-gencode arch=compute_20, code=sm_21 \
-gencode arch=compute_30, code=sm_30 \
-gencode arch=compute_35, code=sm_35 \
-gencode arch=compute_50, code=sm_50 \
-gencode arch=compute_50, code=compute_50 \

[happyharrycn]

GTX 760 only supports CUDA compute capability version 3.5. So you have to delete the last two lines in your CUDA_ARCH.

[wujiyang]

@happyharrycn Finally I found my answer here:https://github.com/rbgirshick/py-faster-rcnn/issues/2
and thank you all the same. By the way, GTX760 has a computer capability version 3.0