Pip install mamba-ssm

[as7438]

After error installing causal_conv1d, I downloaded "causal_conv1d-1.3.0.post1+cu118torch1.12cxx11abiFALSE-cp310-cp310-linux_x86_64.whl" and pip installed it directly as suggested in another issue ticket.

However, afterwards i tried pip install mamba-ssm, and i got an error:

error: subprocess-exited-with-error

× Getting requirements to build wheel did not run successfully.
│ exit code: 1
╰─> [23 lines of output]
/localscratch/as7438/as7438/pip-build-env-jmrxpe19/overlay/lib/python3.10/site-packages/torch/_subclasses/functional_tensor.py:276: UserWarning: Failed to initialize NumPy: No module named 'numpy' (Triggered internally at /pytorch/torch/csrc/utils/tensor_numpy.cpp:81.)
cpu = _conversion_method_template(device=torch.device("cpu"))
: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.

  torch.__version__  = 2.7.0+cu126


  Traceback (most recent call last):

line 143, in get_requires_for_build_wheel
return hook(config_settings)
File "/localscratch/as7438/as7438/pip-build-env-jmrxpe19/overlay/lib/python3.10/site-packages/setuptools/build_meta.py", line 331, in get_requires_for_build_wheel
return self._get_build_requires(config_settings, requirements=[])
File "/localscratch/as7438/as7438/pip-build-env-jmrxpe19/overlay/lib/python3.10/site-packages/setuptools/build_meta.py", line 301, in _get_build_requires
self.run_setup()
File "/localscratch/as7438/as7438/pip-build-env-jmrxpe19/overlay/lib/python3.10/site-packages/setuptools/build_meta.py", line 317, in run_setup
exec(code, locals())
File "", line 188, in
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

× Getting requirements to build wheel did not run successfully.
│ exit code: 1
╰─> See above for output.

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

[dani-capellan]

I ran into similar issues (and some more commented in other threads) with mamba-ssm and causal_conv1d. In your case, seems like you are trying to install it in a node without GPU support/NVCC.

Here is what worked for me:

First, I verified that the following versions were installed:

python==3.10  
torch==2.0.1  
torchvision==0.15.2

Once those versions were in place, I followed these steps:

1. Install mamba-ssm v1.2.0.post1 from wheel (.whl)

• Go to the v1.2.0.post1 release page.
• Download the appropriate wheel for your environment.
  In my case:
  mamba_ssm-1.2.0.post1+cu118torch2.0cxx11abiFALSE-cp310-cp310-linux_x86_64.whl
• Then install it using:

pip install mamba_ssm-1.2.0.post1+cu118torch2.0cxx11abiFALSE-cp310-cp310-linux_x86_64.whl

2. Install causal-conv1d v1.2.0.post1 from wheel (.whl)

• Go to the v1.2.0.post1 release page.
• Download the correct wheel.
  In my case:
  causal_conv1d-1.2.0.post1+cu118torch2.0cxx11abiFALSE-cp310-cp310-linux_x86_64.whl
• Install it with:

pip install causal_conv1d-1.2.0.post1+cu118torch2.0cxx11abiFALSE-cp310-cp310-linux_x86_64.whl

---

Hope this helps! Let me know if you run into any issues.