Blackwell version with cuda 12.8 please!

[dirtmonster1337]

Hello!

Some of us have 50xx cards that use Blackwell/sm_120 architecture and we can't use RVC until this is fixed :(

What we need

-the embedded python/pytorch to use the 'nightly' build from

https://pytorch.org/
pip3 install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu128

Could someone please fork it or fix it using this pytorch?

[sakatatoushirou]

You can first install the nightly version of torch yourself.
First, uninstall the pre-installed version.

.\runtime\python.exe -m pip uninstall torch torchaudio torchvision -y

Then install the torchcu128 version.

.\runtime\python.exe -m pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu128

[dirtmonster1337]

WARNING: The script isympy.exe is installed in .RVC1006Nvidia\runtime\Scripts' which is not on PATH.
Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.
WARNING: The scripts torchfrtrace.exe and torchrun.exe are installed in .\RVC1006Nvidia\runtime\Scripts' which is not on PATH.
Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
torch-directml 0.2.0.dev230426 requires torch==2.0.0, but you have torch 2.8.0.dev20250323+cu128 which is incompatible.
torch-directml 0.2.0.dev230426 requires torchvision==0.15.1, but you have torchvision 0.22.0.dev20250324+cu128 which is incompatible.
xformers 0.0.19 requires torch==2.0.0, but you have torch 2.8.0.dev20250323+cu128 which is incompatible.

Any thoughts on this?

[dirtmonster1337]

on load it says

WARNING | xformers | WARNING[XFORMERS]: xFormers can't load C++/CUDA extensions. xFormers was built for:
PyTorch 2.0.0+cu118 with CUDA 1108 (you have 2.8.0.dev20250323+cu128)
Python 3.9.13 (you have 3.9.13)
Please reinstall xformers (see https://github.com/facebookresearch/xformers#installing-xformers)
Memory-efficient attention, SwiGLU, sparse and more won't be available.
Set XFORMERS_MORE_DETAILS=1 for more details
2025-03-24 09:25:22 | INFO | configs.config | Found GPU NVIDIA GeForce RTX 5080
is_half:True, device:cuda:0