Open tranctan opened 4 years ago
Good question, though I don't know what can be done from the wrapper side. Maybe it has to be done on the C code side (from the original repo)?
Hi, thanks for your ideas! Since there is another issue #55 mentioning about bringing the code to GPU with CuPy. I'm not experienced much with Python wrapping, so do you think this is possible ?
If not then the only way has to be on the C code side.
I was expecting an answer from the community, too. 😅 I don't have experience on CuPy. Not sure how that works, but I thought it's the C++ code in WORLD that does the heavy lifting in PyWORLD. Anyway, please let us know if you try anything out.
Sure, I think I will leave this issue open if there is anyone out there interested !
Good question, though I don't know what can be done from the wrapper side. Maybe it has to be done on the C code side (from the original repo)?
from the wrapper side, please release the GIL each time when running time-consuming c functions, so users can run pyworld multithreaded
Hi, may I ask if you any suggestions for speeding up PyWorld? I have been thinking about this but have not figured out which is a proper way.
Looking forward to your comments. :D