mamba-ssm complaining nvcc not found while cuda is installed

[chengmarc]

Main Complaint

C:\Users\marcc\AppData\Local\Temp\pip-build-env-3vuoamb7\overlay\Lib\site-packages\torch\_subclasses\functional_tensor.py:295:  
UserWarning: Failed to initialize NumPy: No module named 'numpy' (Triggered internally at C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\torch\csrc\utils\tensor_numpy.cpp:84.)
        cpu = _conversion_method_template(device=torch.device("cpu"))

<string>:118: UserWarning: mamba_ssm was requested, but nvcc was not found.  Are you sure your environment has nvcc available?  If you're installing within a container from https://hub.docker.com/r/pytorch/pytorch, only images whose names contain 'devel' will provide nvcc. 

note: I am in a conda environment with the following:

I have causal-conv1d==1.4.0, NumPy is installed, too.

More Info

 Traceback (most recent call last):
        File "C:\Users\marcc\miniconda3\envs\default\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 353, in <module>
          main()
        File "C:\Users\marcc\miniconda3\envs\default\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 335, in main
          json_out['return_val'] = hook(**hook_input['kwargs'])
        File "C:\Users\marcc\miniconda3\envs\default\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 118, in get_requires_for_build_wheel
          return hook(config_settings)
        File "C:\Users\marcc\AppData\Local\Temp\pip-build-env-7smda003\overlay\Lib\site-packages\setuptools\build_meta.py", line 334, in get_requires_for_build_wheel
          return self._get_build_requires(config_settings, requirements=[])
        File "C:\Users\marcc\AppData\Local\Temp\pip-build-env-7smda003\overlay\Lib\site-packages\setuptools\build_meta.py", line 304, in _get_build_requires
          self.run_setup()
        File "C:\Users\marcc\AppData\Local\Temp\pip-build-env-7smda003\overlay\Lib\site-packages\setuptools\build_meta.py", line 320, in run_setup
          exec(code, locals())
        File "<string>", line 188, in <module>
      NameError: name 'bare_metal_version' is not defined
      [end of output]

  note: This error originates from a subprocess, and is likely not a problem with pip.
error: subprocess-exited-with-error

[tridao]

Is nvcc installed?

[chengmarc]

C:\Users\marcc>nvcc --version

gives

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Wed_Feb__8_05:53:42_Coordinated_Universal_Time_2023
Cuda compilation tools, release 12.1, V12.1.66
Build cuda_12.1.r12.1/compiler.32415258_0

[tridao]

Likely setup.py can't find the right path to nvcc. We rely on CUDA_HOME from torch.utils.cpp_extension What does from torch.utils.cpp_extension import CUDA_HOME give?

[chengmarc]

>>> from torch.utils.cpp_extension import CUDA_HOME and CUDA_HOME gives: C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.1

[jq-Qin]

Hello，have you solved your problem？I met it too.

[chengmarc]

this problem hasn't been solved

Hello，have you solved your problem？I met it too.

[So-cean]

same problem

[bishnukhadka]

Anyone solved this issue yet?

[matkara]

Same problem on windows, using venv. CUDA_HOME returns the expected path.

[setday]

I ran into the same error.

As I understand it, the builder couldn't find the right toolы in an isolated environment, so I just added the --no-build-isolation flag and it helped!

So, try to install it with python -m pip install mamba-ssm --no-build-isolation

[zhoujy1321]

I ran into the same error.

As I understand it, the builder couldn't find the right toolы in an isolated environment, so I just added the flag and it helped!--no-build-isolation

So, try to install it with python -m pip install mamba-ssm --no-build-isolation

don't work(╥﹏╥)

[bishnukhadka]

I ran into the same error.

As I understand it, the builder couldn't find the right toolы in an isolated environment, so I just added the --no-build-isolation flag and it helped!

So, try to install it with python -m pip install mamba-ssm --no-build-isolation

For me I used the Windows Subsystem for Linux (WSL) and used the --no-build-isolation. It did install properly. If you are using a Windows system, then consider using WSL.

However, I am not able to use either Mamba1 or Mamba2.

[alarv]

On AWS Sagemaker jupyter lab I had to run conda install -c nvidia cuda-nvcc to install the compatible nvcc compiler with the cuda version I was using, the default image didn't have the right one.