CUDA OOM error

[nautics889]

It seems like something tried to eat all of my memory.

When trying to run via inference_realesrgan.py.
A command to run:

(real-esrgan) PS C:\Users\***\Real-ESRGAN> python .\inference_realesrgan.py -n RealESRGAN_x4plus_anime_6B.pth -i input.jpg -o .\output\

Produced output:

...

100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 17.1M/17.1M [00:01<00:00, 10.7MB/s]
Testing 0 02
Error CUDA out of memory. Tried to allocate 110.72 GiB (GPU 0; 8.00 GiB total capacity; 28.49 GiB already allocated; 0 bytes free; 29.31 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
If you encounter CUDA out of memory, try to set --tile with a smaller number.
Testing 1 03

It works ok via ncnn-vulkan though, so I just keep using it, I'm just curious about this error.