即使设置了 CUDA_HOME 也会出现 CUDA_HOME 错误

Question

我最近尝试运行nvidia labs开发的neuralangelo程序，在设置conda环境时遇到错误：

Collecting git+https://github.com/NVlabs/tiny-cuda-nn/#subdirectory=bindings/torch (from -r C:\Users\shaun\Documents\neuralangelo\requirements.txt (line 3))
  Cloning https://github.com/NVlabs/tiny-cuda-nn/ to c:\users\shaun\appdata\local\temp\pip-req-build-ltveby0h
  Resolved https://github.com/NVlabs/tiny-cuda-nn/ to commit 2ec562e853e6f482b5d09168705205f46358fb39
  Preparing metadata (setup.py): started
  Preparing metadata (setup.py): finished with status 'error'

Pip subprocess error:
  Running command git clone --filter=blob:none --quiet https://github.com/NVlabs/tiny-cuda-nn/ 'C:\Users\shaun\AppData\Local\Temp\pip-req-build-ltveby0h'
  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
  ╰─> [9 lines of output]
      C:\Users\shaun\AppData\Local\Temp\pip-req-build-ltveby0h\bindings/torch\setup.py:5: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/ll
        from pkg_resources import parse_version
      Traceback (most recent call last):
        File "<string>", line 2, in <module>
        File "<pip-setuptools-caller>", line 34, in <module>
        File "C:\Users\shaun\AppData\Local\Temp\pip-req-build-ltveby0h\bindings/torch\setup.py", line 51, in <module>
          raise EnvironmentError("Unknown compute capability. Specify the target compute capabilities in the TCNN_CUDA_ARCHITECTURES environment variable or install PyTorch with the CU)
      OSError: Unknown compute capability. Specify the target compute capabilities in the TCNN_CUDA_ARCHITECTURES environment variable or install PyTorch with the CUDA backend to detec.
      Building PyTorch extension for tiny-cuda-nn version 1.7
      [end of output]

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

× Encountered error while generating package metadata.
╰─> See above for output.

note: This is an issue with the package mentioned above, not pip.
hint: See above for details.

failed

CondaEnvException: Pip failed

我在我的计算机和运行 Neuronangelo 的环境中设置了 CUDA_HOME 变量，但错误不断发生，我不知道为什么。有谁知道这个问题的解决办法吗？

Answer 1

我同意@talonmies。你必须弄清楚架构或 GPU。许多 GPU 代码可以针对架构进行优化，因此您需要设置

TCNN_CUDA_ARCHITECTURES=86

。这会将架构设置为

Ampere 86

。有一个完整的 GPU 和架构列表可供您查找。 查看此页面