Error while installing dependencies

[Cantersoft]

I don't know what CUDA Toolkit is and I don't think I have anything like that installed. Do I need to create an environment variable? Where should that be? I have no technical knowledge in this field so I would appreciate a detailed explanation of steps.

OS: Windows 10 22H2
CPU: R9 7950X
GPU: RTX 3090
RAM: 64GB

Collecting flash_attn (from -r requirements/framework.txt (line 16))
Using cached flash_attn-2.8.3.tar.gz (8.4 MB)
Preparing metadata (setup.py) ... error
error: subprocess-exited-with-error

× python setup.py egg_info did not run successfully.
│ exit code: 1
╰─> [21 lines of output]
C:\Users\user\AppData\Local\Temp\pip-install-eqykcv8s\flash-attn_32bc8a90ea714de9aa14f880e2c44142\setup.py:106: UserWarning: flash_attn 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.
warnings.warn(
Traceback (most recent call last):
File "", line 2, in
File "", line 34, in
File "C:\Users\user\AppData\Local\Temp\pip-install-eqykcv8s\flash-attn_32bc8a90ea714de9aa14f880e2c44142\setup.py", line 227, in
CUDAExtension(
File "C:\Program Files\Python311\Lib\site-packages\torch\utils\cpp_extension.py", line 1078, in CUDAExtension
library_dirs += library_paths(cuda=True)
^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Program Files\Python311\Lib\site-packages\torch\utils\cpp_extension.py", line 1216, in library_paths
paths.append(_join_cuda_home(lib_dir))
^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Program Files\Python311\Lib\site-packages\torch\utils\cpp_extension.py", line 2416, in _join_cuda_home
raise OSError('CUDA_HOME environment variable is not set. '
OSError: CUDA_HOME environment variable is not set. Please set it to your CUDA install root.

  torch.__version__  = 2.5.1+cu118


  [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.

[Cantersoft]

It looks like the toolkit has to be installed and some environment and path vars have to be manually set. And on top of it, none of the dependencies install correctly. I guess I'll have to figure out how to create a virtual environment to get the correct versions of all the modules.

[Cantersoft]

matplotlib was also needed to run the gradio, which was not in the requirements text

[Cantersoft]

After fixing all of that, I'm seeing "ImportError: cannot import name 'models' from partially initialized module 'vace' (most likely due to a circular import) (C:\VACE\vace_init_.py)" when trying to run python vace/gradios/vace_preprocess_demo.py

[Cantersoft]

Okay, that error was because I had VACE-Annotators in the wrong models folder. I moved it to the models folder in the parent. But now there's another error. UnboundLocalError: cannot access local variable 'RAFT' where it is not associated with a value

[Cantersoft]

I can't find a version of Torch and CUDA that work together. There's just too many problems with this. I have no idea how anyone's managed to get it to work.