GPU support

[EC2EZ4RD]

Because I used chainer and chainerrl, so cupy is needed.
But when I built image, it showed me that

**************************************************
*** WARNING: nvcc not in path.
*** WARNING: Please set path to nvcc.
**************************************************

-------- Configuring Module: cuda --------
command 'gcc' failed with exit status 1

************************************************************
* CuPy Configuration Summary                               *
************************************************************

Build Environment:
  Include directories: []
  Library directories: ['/usr/local/cuda/lib64']
  nvcc command       : (not found)

Environment Variables:
  CFLAGS          : (none)
  LDFLAGS         : (none)
  LIBRARY_PATH    : (none)
  CUDA_PATH       : (none)
  NVTOOLSEXT_PATH : (none)
  NVCC            : (none)

Modules:
  cuda      : No
    -> Include files not found: ['cublas_v2.h', 'cuda.h', 'cuda_profiler_api.h', 'cuda_runtime.h', 'cufft.h', 'curand.h', 'cusparse.h', 'nvrtc.h']
    -> Check your CFLAGS environment variable.

ERROR: CUDA could not be found on your system.
Please refer to the Installation Guide for details:
https://docs-cupy.chainer.org/en/stable/install.html

************************************************************

Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/tmp/pip-install-cflxwwqf/cupy/setup.py", line 64, in <module>
    ext_modules = cupy_setup_build.get_ext_modules()
  File "/tmp/pip-install-cflxwwqf/cupy/cupy_setup_build.py", line 583, in get_ext_modules
    extensions = make_extensions(arg_options, compiler, use_cython)
  File "/tmp/pip-install-cflxwwqf/cupy/cupy_setup_build.py", line 379, in make_extensions
    raise Exception('Your CUDA environment is invalid. '
Exception: Your CUDA environment is invalid. Please check above error log.

So, could you please help to solve it?
Many thanks

[EC2EZ4RD]

And I found this docker used cuda:9.0-runtime-ubuntu16.04 as base docker which is without nvcc.
So I tried add apt.txt with:

apt-utils
libcudnn7-dev=$CUDNN_VERSION-1+cuda9.0

But it still did not work with warning: "debconf: delaying package configuration, since apt-utils is not installed".

[spMohanty]

Hi, You can also include a Dockerfile at the root of your repository, and it will be used instead to build the image. Cheers, Mohanty

…

On Thu, Oct 4, 2018 at 10:55 AM EC2EZ4RD ***@***.***> wrote: And I found this docker used cuda:9.0-runtime-ubuntu16.04 as base docker which is without nvcc. So I tried add apt.txt with: apt-utils libcudnn7-dev=$CUDNN_VERSION-1+cuda9.0 But it still did not work with warning: "debconf: delaying package configuration, since apt-utils is not installed". — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#3 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABghACGp5F8K7kM_msvxJ9Fgg6caYFTDks5uhekqgaJpZM4XHmUF> .

[EC2EZ4RD]

Yeah, I am writing a new dockerfile and having a try.