OOM crash on 7b_int8
#7
by KottCh - opened
instead of 3b, I tried seedvr2_7b_int8_convrot.safetensors, and it OOM crashed.
[INFO] Requested to load VideoAutoencoderKLWrapper
[INFO] 0 models unloaded.
[INFO] loaded completely; 478.07 MB loaded, full load: True
[INFO] Requested to load SeedVR2
[INFO] 0 models unloaded.
[INFO] Model SeedVR2 prepared for dynamic VRAM loading. 7944MB Staged. 0 patches attached. Force pre-loaded 576 weights: 3853 KB.
0%| | 0/1 [00:00<?, ?it/s, Model Initializing ... ]
[ERROR] !!! Exception during processing !!! Allocation on device 0 would exceed allowed memory. (out of memory)
Currently allocated : 13.19 GiB
Requested : 999.00 MiB
Device limit : 15.51 GiB
Free (according to CUDA): 25.50 MiB
PyTorch limit (set by user-supplied memory fraction)
: 17179869184.00 GiB
other one
torch.OutOfMemoryError: Allocation on device 0 would exceed allowed memory. (out of memory)
Currently allocated : 9.52 GiB
Requested : 11.46 GiB
Device limit : 15.51 GiB
Free (according to CUDA): 33.50 MiB
PyTorch limit (set by user-supplied memory fraction)
: 17179869184.00 GiB
so if I reduce to few seconds, crash will be avoided? heh, non A practical thing.
KottCh changed discussion status to closed