Compatible with RTX 50

[kimutaku76]

I tried to get an RTX 5090 and continue RVC work, but It seems that there is a conflict because sm_120 is not supported.

When running RVC, the gpu output is true, but voice synthesis fails and stops in pause.

E:\RVC1006Nvidia>runtime\python.exe infer-web.py --pycmd runtime\python.exe --port 7897 E:\RVC1006Nvidia\runtime\lib\site-packages\torch\cuda_init_.py:173: UserWarning: NVIDIA GeForce RTX 5090 with CUDA capability sm_120 is not compatible with the current PyTorch installation. The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_61 sm_70 sm_75 sm_80 sm_86 sm_90 compute_37. If you want to use the NVIDIA GeForce RTX 5090 GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/

warnings.warn(incompatible_device_warn.format(device_name, capability, " ".join(arch_list), device_name)) 2025-02-13 17:42:56 | INFO | configs.config | Found GPU NVIDIA GeForce RTX 5090 is_half:True, device:cuda:0

I tried to somehow make it compatible with pytorch, but I've spent over 40 hours trying to solve it on my own, but I can't solve it because I'm not an expert.

Can you please upgrade the version?

[syd2333]

I encountered and solved a similar issue. While my solution was for a different AI voice project, the underlying CUDA compatibility problem is the same :

https://github.com/RVC-Boss/GPT-SoVITS/issues/2026

[kimutaku76]

Thank you. First, let's implement it according to your suggestion. :)

[jmeboypc]

Will RVC be updated once the latest PyTorch version releases on Windows?

[DEUPLECT]

Did it fix your problem as suggested by @syd2333 ?

[dirtmonster1337]

running into same problems with sm_120 did @syd2333 work for RVC project?