crash: RVC1006Nvidia Real-Time

when i click start with pre models in RVC it crashing
Real-Time

runtime\python.exe gui_v1.py
is_half:True, device:cuda:0
Input device: 1:Microphone (Realtek(R) Audio) (MME)
Output device: 19:Speakers (Realtek(R) Audio) (Windows DirectSound)
cuda_is_available: True
Traceback (most recent call last):
File "C:\Users\ddr7\Desktop\voice\RVC1006Nvidia_2.7\RVC1006Nvidia\tools\rvc_for_realtime.py", line 106, in init
hubert_model = hubert_model.half()
File "C:\Users\ddr7\Desktop\voice\RVC1006Nvidia_2.7\RVC1006Nvidia\runtime\lib\site-packages\torch\nn\modules\module.py", line 1001, in half
return self._apply(lambda t: t.half() if t.is_floating_point() else t)
File "C:\Users\ddr7\Desktop\voice\RVC1006Nvidia_2.7\RVC1006Nvidia\runtime\lib\site-packages\torch\nn\modules\module.py", line 797, in _apply
module._apply(fn)
File "C:\Users\ddr7\Desktop\voice\RVC1006Nvidia_2.7\RVC1006Nvidia\runtime\lib\site-packages\torch\nn\modules\module.py", line 797, in _apply
module._apply(fn)
File "C:\Users\ddr7\Desktop\voice\RVC1006Nvidia_2.7\RVC1006Nvidia\runtime\lib\site-packages\torch\nn\modules\module.py", line 797, in _apply
module._apply(fn)
[Previous line repeated 1 more time]
File "C:\Users\ddr7\Desktop\voice\RVC1006Nvidia_2.7\RVC1006Nvidia\runtime\lib\site-packages\torch\nn\modules\module.py", line 820, in _apply
param_applied = fn(param)
File "C:\Users\ddr7\Desktop\voice\RVC1006Nvidia_2.7\RVC1006Nvidia\runtime\lib\site-packages\torch\nn\modules\module.py", line 1001, in
return self._apply(lambda t: t.half() if t.is_floating_point() else t)
RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

Start with tools/rvc_for_realtime.py at line 106, where hubert_model.half() raises the reported CUDA error. Reproduce the crash by clicking Start with pretrained models and review the bundled runtime's PyTorch/CUDA compatibility. Done means real-time conversion starts successfully on the reported NVIDIA setup without this error.