CUDA Out of Memory Error on Base Install

[carsonbird]

I'm running on Paperspace, Gradient Notebooks, P4000 GPU, and 50GB of persistent storage. I run a default notebook, go through the installation steps, start the gradio page, and keep getting this error whenever I try to run a "prediction" (any kind of ML generation):

OutOfMemoryError: CUDA out of memory. Tried to allocate 10.00 MiB (GPU 0; 7.92 GiB total capacity; 6.80 GiB already allocated; 1.69 MiB free; 7.04 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

Any ideas? Is it just a GPU with too low internal memory? I see that stable diffusion's recommendation is a GPU with at least 6GB VRAM, and this one has 8GB, so why would that be the problem?

[TheLastBen]

If your running SDXL, 8GB might not always work, it might work for 1024x1024, but if you do high res or refiner, it will crash