Open tna1syv opened 1 month ago
I am assuming you are training your own dataset. I faced the same problem. Turns out it was because my images were too high quality and big (and in png format). I recommend downsizing your data (maybe to 960x540 pixels) and giving it in jpg format rather than png format. Png format takes up more space than jpg.
@nurbanu170399 Sorry I can't help you without detailed errors. I suggest you debug the code and cast the related arrays to numpy.array.
Thank you for the great work and sharing the code!
I run into a
cuda out of memory
error while training. I'm using V100 with 16GB memory GPU. I'm wondering if there is any work around?Thanks!