search

saifhassan / Wav2Lip-HD

High-Fidelity Lip-Syncing with Wav2Lip and Real-ESRGAN
Other
320 stars 75 forks source link

CUDA out of memory, GPU up But "Utilization" = LOW, worked yesterday and now it does not work! #32

Open Stable777 opened 8 months ago

Stable777 commented 8 months ago

Hi, got this problem:

torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 15.82 GiB. GPU 0 has a total capacty of 23.99 GiB of which 0 bytes is free. Of the allocated memory 21.37 GiB is allocated by PyTorch, and 13.42 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

Wav2lip-HD worked ysterday, but it does not today, it says out of memory at the UPSCALING step with REAL ersgan, I tried another repo and noticed that the GPU is as UP as 24 G , but in utilization it is slow image

Any help

tranthuhoai3786 commented 2 months ago

I have the same problem. Do you have any solution? @Stable777 @saifhassan