How can we run this on a 50 series GPU

[AswinPG]

The instructions are for cuda 11.8, which is not supported by RTX 50 series. Is it possible to use it on 50 series gpu's? :)

[Pr3zLy]

use cudnn8.5.0.96_cuda11
add pygrabber to requirements.txt
delete venv folder, reinstall requirements.txt
run the command: pip install torch==2.3.1 torchvision==0.18.1 torchaudio==2.3.1 --index-url https://download.pytorch.org/whl/cu118
now should work

[AswinPG]

With CUDA 11.8 right ?

[AswinPG]

I am getting this error after trying with the above configuration
2025-06-08 18:26:57.0461826 [E:onnxruntime:, sequential_executor.cc:514 onnxruntime::ExecuteKernel] Non-zero status code returned while running Gemm node. Name:'fc1' Status Message: D:\a_work\1\s\onnxruntime\core\providers\cuda\cuda_call.cc:121 onnxruntime::CudaCall D:\a_work\1\s\onnxruntime\core\providers\cuda\cuda_call.cc:114 onnxruntime::CudaCall CUBLAS failure 3: CUBLAS_STATUS_ALLOC_FAILED ; GPU=0 ; hostname=SUNDAY ; file=D:\a_work\1\s\onnxruntime\core\providers\cuda\cuda_execution_provider.cc ; line=168 ; expr=cublasCreate(&cublas_handle_);

[AswinPG]

--extra-index-url https://download.pytorch.org/whl/cu126

numpy>=1.23.5,<2
typing-extensions>=4.8.0
opencv-python==4.10.0.84
cv2_enumerate_cameras==1.1.15
onnx==1.16.0
insightface==0.7.3
psutil==5.9.8
tk==0.1.0
customtkinter==5.2.2
pillow==11.1.0

CUDA 12.6 build for PyTorch and TorchVision

torch==2.7.1+cu126; sys_platform != 'darwin'
torch==2.7.1; sys_platform == 'darwin'

torchvision
torchvision==0.18.0; sys_platform == 'darwin'

ONNX Runtime: updated for CUDA 12.6 compatibility

onnxruntime-silicon==1.16.3; sys_platform == 'darwin' and platform_machine == 'arm64'
onnxruntime-gpu==1.20.0; sys_platform != 'darwin'

TensorFlow - ensure the version installed supports CUDA 12.6 (e.g., TF 2.16+)

tensorflow; sys_platform != 'darwin'

opennsfw2==0.10.2
protobuf==4.23.2
pygrabber

This configuration works with CUDA 12.6 and cudnn 9.0

Note : TensorRT is still not working