dire flame
when trying to launch the cuda version, everything works fine, but when trying to install the dml version i get this : "pyimod03_ctypes.install..PyInstallerImportError: Failed to load dynlib/dll 'C:\Users[username]\Desktop\dist\main_internal\torch_directml\DirectML.dll'. Most likely this dynlib/dll was not found when the application was frozen."
and i use an amd rx 9060XT so obviously the cuda version doesn't recognize my gpu
any idea how to fix it ?
or should i just download something else ?
high oyster
when trying to launch the cuda version, everything works fine, but when trying t...
Are you trying to run the original version of W-Okada or something? The original version of W-Okada DirectML is bugged. I use the better one, which is Deitertis' fork W-Okada.
dire flame
This is the better one. https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w...
oh thanks, what's the difference / what's better with this fork ?
high oyster
oh thanks, what's the difference / what's better with this fork ?
The difference is mostly performance. The DirectML (AMD/Intel GPU) variant is better on fork W-Okada, while the original one is bugged.