How much GPU memory does the x-part need? #9
Description
torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 9.26 GiB. GPU 0 has a total capacity of 95.00 GiB of which 5.02 GiB is free. Process 95489 has 13.56 GiB memory in use. Process 295058 has 2.35 GiB memory in use. Including non-PyTorch memory, this process has 74.05 GiB memory in use. Of the allocated memory 69.75 GiB is allocated by PyTorch, and 3.87 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)
I am using an H20 with 96GB GPU, and try to process a low-poly glb file through x-part, but meet this oom problem. 😭
How much GPU memory does the x-part need for inference?