sepconv
sepconv copied to clipboard
when complier the sepcon,occur the following error?
src/SeparableConvolution_kernel.cu(66): error: pointer to incomplete class type is not allowed ... src/SeparableConvolution_kernel.cu(170): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(170): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(171): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(171): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(173): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(173): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(174): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(174): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(176): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(176): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(177): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(177): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(179): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(179): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(180): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(180): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(181): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(181): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(182): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(182): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(189): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(189): error: pointer to incomplete class type is not allowed
src/SeparableConvolution_kernel.cu(190): error: pointer to incomplete class type is not allowed
it like the it doesnot know the THCudaTensor,any ideas? THCudaTensor* input, THCudaTensor* vertical, THCudaTensor* horizontal, THCudaTensor* output
must use torch=0.4.0 0.4.1 won't work, seems to due to https://github.com/pytorch/pytorch/commit/4caea64d729728e3304e9c1e97081a6bdd463913#diff-62544837c1a693cd4048c2b56f10a968
remember to use python 3.6, so u can pip install torch=0.4.0 0.4.0 not available if python 3.7
must use torch=0.4.0 0.4.1 won't work, seems to due to pytorch/pytorch@4caea64#diff-62544837c1a693cd4048c2b56f10a968
remember to use python 3.6, so u can pip install torch=0.4.0 0.4.0 not available if python 3.7
must use torch=0.4.0 0.4.1 won't work, seems to due to pytorch/pytorch@4caea64#diff-62544837c1a693cd4048c2b56f10a968
remember to use python 3.6, so u can pip install torch=0.4.0 0.4.0 not available if python 3.7
hello,I am a student from Huazhong University of Science and Technology. And i want to use the code to train the model, but the compile is always wrong. Could you please briefly explain the compilation environment, and compile steps? Thank you so much!
best wishes
what is your python version and torch version? i went smoothly with the read me doc, as long as the torch= 0.4.0, python = 3.6
i use anaconda on ubuntu 16.04, but whether u use anaconda or virtual env, should not have big difference
I got this error:"ImportError: No module named 'libs.sepconv._ext' ",this error derives from "./libs/sepconv/SeparableConvolution.py" ,its import part said "import libs.sepconv._ext as _ext" .However,I didn't find lib.sepconv._ext either. My env is ubuntu 16.04,python3.5 and torch1.1.0,could you help me?
Try get torch= 0.4.0, python = 3.6. it will be a lot easier
Try get torch= 0.4.0, python = 3.6. it will be a lot easier
Hello Eric! According to what you said, I have run the program successfully, but the training isn't working well. The parameters I used are set according to your paper, a batch size of 16 and a learning rate of 0.01, epochs =1000 or 2000, but the result is still blur. Could you please give me some tips about the training process?
Try get torch= 0.4.0, python = 3.6. it will be a lot easier
Hi, I have installed everything according to the requirements however, I still get this ModuleNotFoundError: No module named 'libs.sepconv._ext.cunnex._cunnex' Do you have any ideas on how to solve this? Thanks
@chenwu-mc Have you solved the blur problem finally?