using too much vram

Open Uemuet opened this issue 1 year ago • 0 comments

lately i get this failure that im going to the limits of my vram and i dont know why, because i can get only render less frame batches than before and get this.

torch.cuda.OutOfMemoryError: Allocation on device 0 would exceed allowed memory. (out of memory) Currently allocated : 27.28 GiB Requested : 22.21 GiB Device limit : 39.56 GiB Free (according to CUDA): 20.81 MiB PyTorch limit (set by user-supplied memory fraction) : 17179869184.00 GiB

and a month before i could render 400 frames in a high resolution. My environment is colab pro

May 11 '24 06:05 Uemuet