search

endernewton / tf-faster-rcnn

Tensorflow Faster RCNN for Object Detection
https://arxiv.org/pdf/1702.02138.pdf
MIT License
3.65k stars 1.58k forks source link

OSError: The nvcc binary could not be located in your $PATH. Either add it to your path, or set $CUDAHOME #391

Open 1451595897 opened 5 years ago

1451595897 commented 5 years ago

If I make in the lib folder(in google colab),it shows that “OSError: The nvcc binary could not be located in your $PATH. Either add it to your path, or set $CUDAHOME.”I don't know how to deal with this problem? image

rnsandeep commented 5 years ago

You need to add the cuda path to bashrc. export CUDA_HOME=/usr/local/cuda-8.0 export LD_LIBRARY_PATH=${CUDA_HOME}/lib64

PATH=${CUDA_HOME}/bin:${PATH} export PATH

1451595897 commented 5 years ago

However,I add the cuda path to bashrc later,this problem still exists. @rnsandeep image

rnsandeep commented 5 years ago

You can look at this thread. https://github.com/rbgirshick/py-faster-rcnn/issues/123

1451595897 commented 5 years ago

Thank you. I can use CPU to run this demo in google colab when I modify the setup.py and the nms_wrapper.py. But I still want to use GPU. I am unable to fix it. @rnsandeep

rnsandeep commented 5 years ago

For working with gpu you need to install nvidia driver and cuda properly. and then do make. Please let me know which gpu you are using and linux environment so that i can refer which version to install.

LP308210365 commented 5 years ago

Have you solved this problem? i meet the same problem ,could you please tell me how to sovle it

psychiatrist-xsd commented 5 years ago

I also meet this problem, just because i install the tensorflow in the virtual environment by the command “conda install tensorflow-gpu=×** ",and it also install part of cuda （cudatoolkit) corresponding to my tensorflow。That means there is not nvcc in my computer （I never install cuda in my computer ) ， so i meet the problem above, to solve it i install the complete cuda ,and set the PATH .just OK~

LP308210365 commented 5 years ago

thank you very much. i will try it later

---Original--- From: "shidunxu"<notifications@github.com> Date: Fri, Aug 30, 2019 10:16 AM To: "endernewton/tf-faster-rcnn"<tf-faster-rcnn@noreply.github.com>; Cc: "Comment"<comment@noreply.github.com>;"LP308210365"<308210365@qq.com>; Subject: Re: [endernewton/tf-faster-rcnn] OSError: The nvcc binary could not be located in your $PATH. Either add it to your path, or set $CUDAHOME (#391)

I also meet this problem, just because i install the tensorflow in the virtual environment by the command “conda install tensorflow-gpu=×** ",and it also install part of cuda （cudatoolkit) corresponding to my tensorflow。That means there is not nvcc in my computer （I never install cuda in my computer ) ， so i meet the problem above, to solve it i install the complete cuda ,and set the PATH .just OK~

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

zhangfengjiao-zju commented 3 years ago

I also meet this problem, just because i install the tensorflow in the virtual environment by the command “conda install tensorflow-gpu=×** ",and it also install part of cuda （cudatoolkit) corresponding to my tensorflow。That means there is not nvcc in my computer （I never install cuda in my computer ) ， so i meet the problem above, to solve it i install the complete cuda ,and set the PATH .just OK~

is your GPU from either intel or Nvidia. My GPU is from Intel, can I istall cuda?

federiconuta commented 3 years ago

@zhangfengjiao-zju I think you cannot. I have the same issue as you do