Skip to content

CUDA out of memory #106

@meera

Description

@meera

Running inference_propainter.py

Works great for samples - but trying to paint 5 second long video and getting this error.
subvideo_length is 2

torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 7.82 GiB. GPU 0 has a total capacity of 23.60 GiB of which 7.81 GiB is free. Including non-PyTorch memory, this process has 15.78 GiB memory in use. Of the allocated memory 14.89 GiB is allocated by PyTorch, and 596.41 MiB 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)

 export PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True
python inference_propainter.py --video myvideo_frames --mask myvideo_mask/ --subvideo_length  2

Any suggestions?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions