princeton-vl / Coupled-Iterative-Refinement

MIT License
104 stars 20 forks source link

environment creation fails for lietorch #8

Open JKBehrens opened 1 year ago

JKBehrens commented 1 year ago

Hi,

I tried to install your code, and conda returned this error. I guess I can clone and install lietorch myself, but I wanted to leave this not for you.

Best, Jan

Pip subprocess error:
  Running command git clone --quiet https://github.com/princeton-vl/lietorch.git /tmp/pip-req-build-u11m6szo
  Running command git submodule update --init --recursive -q
  error: subprocess-exited-with-error

  × python setup.py egg_info did not run successfully.
  │ exit code: 1
  ╰─> [12 lines of output]
      Traceback (most recent call last):
        File "<string>", line 2, in <module>
        File "<pip-setuptools-caller>", line 34, in <module>
        File "/tmp/pip-req-build-u11m6szo/setup.py", line 16, in <module>
          CUDAExtension('lietorch_backends',
        File "/home/behrejan/.conda/envs/mybase/envs/cir/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 867, in CUDAExtension
          library_dirs += library_paths(cuda=True)
        File "/home/behrejan/.conda/envs/mybase/envs/cir/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 975, in library_paths
          if (not os.path.exists(_join_cuda_home(lib_dir)) and
        File "/home/behrejan/.conda/envs/mybase/envs/cir/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1982, in _join_cuda_home
          raise EnvironmentError('CUDA_HOME environment variable is not set. '
      OSError: CUDA_HOME environment variable is not set. Please set it to your CUDA install root.
      [end of output]
francescosarno commented 1 year ago

I have the same issue, did you manage to solve?

mqtjean commented 1 year ago

Hi,

The error here is that your CUDA_HOME environment is not set.

Do which nvcc

if you don't have nvcc, install cudatoolkit with the same version you build pytorch with

then redo which nvcc

Then if the path is for example : /home/usr/anaconda/env_name/bin/nvcc

Do : export CUDA_HOME=/home/usr/anaconda/env_name/ the path without bin/nvcc